<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[COMPFEST - Medium]]></title>
        <description><![CDATA[COMPFEST is the BIGGEST student-held IT Event arranged by the students of Faculty of Computer Science, University of Indonesia. - Medium]]></description>
        <link>https://medium.com/compfest?source=rss----8984eec5fa5d---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>COMPFEST - Medium</title>
            <link>https://medium.com/compfest?source=rss----8984eec5fa5d---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 07:13:38 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/compfest" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Breaking the Hacking Stereotype: Discovering the Essence of System Protection at XCelerate Batch 2…]]></title>
            <link>https://medium.com/compfest/breaking-the-hacking-stereotype-discovering-the-essence-of-system-protection-at-xcelerate-batch-2-6432d0f922d9?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/6432d0f922d9</guid>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[university]]></category>
            <dc:creator><![CDATA[Malvin]]></dc:creator>
            <pubDate>Mon, 14 Sep 2026 13:50:56 GMT</pubDate>
            <atom:updated>2026-09-14T13:50:54.477Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Breaking the Hacking Stereotype: Discovering the Essence of System Protection at XCelerate Batch 2 COMPFEST 18</strong></h3><p><a href="https://medium.com/@malvinlionard/31b8378099a6">Baca Versi Bahasa Indonesia di Sini</a></p><p>Amid the rapid digitalization of banking services and public infrastructure, the scope of cyber threats is evolving at an unprecedented pace. The rising complexity of digital attacks creates an urgent demand for cybersecurity practitioners. Nevertheless, a digital skills gap persists in this domain. On one hand, industries urgently require talent capable of designing and maintaining infrastructure at scale. On the other hand, many young people remain anchored to the stereotype that cybersecurity revolves solely around hacking. Addressing this challenge, COMPFEST 18 provided a comprehensive learning space through the XCelerate Batch 2 seminar titled “Securing the Future: Cybersecurity Skills in Practice.” Held virtually on Sunday, September 6, 2026, XCelerate Batch 2 aimed to equip young talents with the essential knowledge needed to navigate system security dynamics across the industry.</p><p>The seminar represents a tangible step in realizing COMPFEST 18’s overarching theme: “Driving Infinite Growth to Enclose the Trench of Digital Skill Gap.” The enthusiasm of hundreds of university students across the Zoom session reflected strong interest in the cybersecurity landscape. As part of the XCelerate series, the seminar was designed to bridge student curiosity with the technical demands of today’s industry.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-y47kiUJ0Aic2xBW3nE_HQ.png" /><figcaption>The opening session of the COMPFEST 18 XCelerate Batch 2 Seminar, “Securing the Future: Cybersecurity Skills in Practice,” held via Zoom Meeting. (Source: Documentation)</figcaption></figure><p>Interestingly, the day-to-day reality of cybersecurity diverges significantly from popular portrayals on the internet and social media. Keynote speaker Dean Verrell Matalino Ariefin, Senior Information Technology Network Specialist at BCA, prompted participants to reflect on the true meaning of information security. Drawing from audience responses, Verrell clarified that enterprise cybersecurity fundamentally centers on system protection, infrastructure integrity, and the resilience of network defense mechanisms. The ability to identify threats — ranging from Trojan risks to modern cyberattacks — serves as the foundational baseline required before designing advanced security systems.</p><p>To ensure participants went beyond theoretical concepts, XCelerate Batch 2 challenged them to hone their analytical thinking directly. Verrell led attendees into an interactive investigation simulation using a dedicated mini CTF (Capture The Flag) platform hosted at xcelerate-mini-ctf. This hands-on scenario challenged participants to think critically, identify security loopholes, and adopt the defensive mindset essential for vulnerability analysis. The approach not only heightened engagement but also illustrated that cybersecurity operates as an art of high-precision problem-solving.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7d0gCxI38n9-yWR7cL3tbg.png" /><figcaption><em>Practical cybersecurity presentation by Dean Verrell Matalino Ariefin (Senior IT Network Specialist at BCA). (Source: Documentation)</em></figcaption></figure><p>While the technical challenge showcased vulnerability discovery, the subsequent Q&amp;A session — moderated by Ilma Alpha Mannix, S.Kom., M.Kom., lecturer and researcher at Fasilkom UI — highlighted how these principles apply to real-world banking operations. Verrell underscored the marked distinction between lab-scale security experiments and managing thousands of network devices across a national banking infrastructure. In enterprise environments, security remains inextricably tied to robust network architecture design. He emphasized mastering end-to-end design documentation, ranging from High-Level Design (HLD) for mapping macro-level architecture and system topology to Low-Level Design (LLD) detailing granular configurations, ports, and data flow rules.</p><p>Complementing these architectural insights, Verrell outlined a roadmap for students aspiring to cultivate industry-standard competencies. He underscored the importance of standardized network security certifications, such as Fortinet NSE (Network Security Expert), to master perimeter defense, firewall administration, and enterprise-grade integrated security. A profound understanding of data communication flows and network configurations is vital, distinguishing seasoned practitioners from those who merely run automated tools without grasping the underlying logic.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fIQWTRXSmFUTjfM4esSOog.png" /><figcaption><em>Interactive talk show and Q&amp;A session with Dean Verrell Matalino Ariefin, moderated by Ilma Alpha Mannix, S.Kom., M.Kom. (Source: Documentation)</em></figcaption></figure><p>The depth of insights unpacked throughout XCelerate Batch 2 marks a vital milestone in COMPFEST 18’s talent development ecosystem. Through the XCelerate series, combined with upcoming hands-on workshops alongside cybersecurity consultants, COMPFEST 18 continuously strives to equip young talents with solid fundamentals and practical skills. These initiatives complement the upcoming national-level CTF competition and dedicated academy tracks leading toward the Main Event.</p><p>For young talents aspiring to stand at the forefront of safeguarding Indonesia’s digital ecosystem, the gateway is wide open. Follow the latest schedules and upcoming program lineups via COMPFEST’s official website and social media channels to take your part in today’s technological transformation. Prepare yourself, sharpen your potential, and enclose the gap!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6432d0f922d9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/breaking-the-hacking-stereotype-discovering-the-essence-of-system-protection-at-xcelerate-batch-2-6432d0f922d9">Breaking the Hacking Stereotype: Discovering the Essence of System Protection at XCelerate Batch 2…</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Mendobrak Stereotip Peretasan: Menemukan Esensi Proteksi Sistem di XCelerate Batch 2 COMPFEST 18]]></title>
            <link>https://medium.com/compfest/mendobrak-stereotip-peretasan-menemukan-esensi-proteksi-sistem-di-xcelerate-batch-2-compfest-18-31b8378099a6?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/31b8378099a6</guid>
            <category><![CDATA[education]]></category>
            <category><![CDATA[university]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <dc:creator><![CDATA[Malvin]]></dc:creator>
            <pubDate>Mon, 14 Sep 2026 13:50:48 GMT</pubDate>
            <atom:updated>2026-09-14T13:50:46.608Z</atom:updated>
            <content:encoded><![CDATA[<p><a href="https://medium.com/compfest/6432d0f922d9">Read the English Version Here</a></p><p>Di tengah pesatnya digitalisasi layanan perbankan dan infrastruktur umum, <em>scope</em> dari ancaman siber sekarang bergerak dengan sangat cepat. Kompleksitas atau kerumitan dari serangan digital yang terus meningkat menciptakan kebutuhan darurat akan praktisi <em>cybersecurity</em>. Walaupun begitu, masih terdapat <em>digital skill gap</em> di area ini. Di satu sisi, industri mendesak adanya mereka yang mampu merancang dan merawat infrastruktur dalam skala yang besar. Di sisi lain, masih banyak dari generasi muda yang terpaku pada stereotip bahwa <em>cybersecurity</em> hanya berfokus pada peretasan. Berangkat dari masalah ini, COMPFEST 18 hadir membawa jawaban dalam bentuk ruang pembekalan yang mendalam melalui Seminar XCelerate Batch 2 berjudul “Securing the Future: Cybersecurity Skills in Practice”. Diselenggarakan secara daring pada Minggu, 6 September 2026, XCelerate Batch 2 hadir untuk membekali talenta muda dengan pengetahuan yang dibutuhkan untuk menghadapi dinamika keamanan sistem di dunia industri ini.</p><p>Hadirnya Seminar XCelerate Batch 2 menjadi wujud nyata komitmen COMPFEST 18 dalam merealisasikan tema besarnya, yaitu “Driving Infinite Growth to Enclose the Trench of Digital Skill Gap”. Antusiasme ratusan peserta dari berbagai perguruan tinggi pada Zoom seminar mencerminkan tingginya minat generasi muda untuk memahami dunia dari <em>cybersecurity</em>. Sebagai bagian dari program XCelerate, seminar ini dirancang untuk menjadi penghubung antara rasa ingin tahu mahasiswa dengan kebutuhan teknis yang dibutuhkan industri saat ini.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-y47kiUJ0Aic2xBW3nE_HQ.png" /><figcaption>Sesi pembukaan Seminar XCelerate Batch 2 COMPFEST 18 bertajuk “Securing the Future: Cybersecurity Skills in Practice” melalui Zoom Meeting. (Sumber: Dokumentasi)</figcaption></figure><p>Menariknya, dunia <em>cybersecurity</em> ternyata memiliki kenyataan yang jauh berbeda dari apa yang kerap ditunjukkan di internet atau media sosial. Topik ini dikulik ketika narasumber utama, Kak Dean Verrell Matalino Ariefin, Senior Information Technology Network Specialist di BCA, mengajak para peserta merefleksikan arti dari keamanan informasi. Dari jawaban-jawaban audiens, Kak Verrell meluruskan pemahaman bahwa esensi dari keamanan siber pada perusahaan ada pada proteksi sistem, integritas infrastruktur, dan ketangguhan mekanisme pertahanan jaringan. Kemampuan mengidentifikasi ancaman, mulai dari bahaya <em>Trojan</em> sampai ke serangan siber yang modern, menjadi fondasi dasar yang harus dipahami sebelum kita dapat membangun sistem keamanan yang canggih.</p><p>Untuk memastikan agar peserta tidak hanya memahami konsep teoritis, Seminar XCelerate Batch 2 ini mengajak para peserta untuk melatih ketajaman logika mereka secara langsung. Kak Verrell mengajak mereka untuk terjun langsung ke simulasi investigasi lewat platform CTF (<em>Capture The Flag</em>) mini yang sudah disiapkan secara khusus lewat platform <em>xcelerate</em>-<em>mini</em>-<em>ctf</em>. Adanya skenario interaktif ini menantang para peserta untuk berpikir kritis, mengenali celah-celah keamanan, dan memahami pola pikir defensif yang diperlukan untuk menganalisis kerentanan sistem. Pendekatan ini tidak hanya meningkatkan antusiasme partisipan, tetapi juga memberikan gambaran bahwa keamanan siber bisa lebih diartikan sebagai seni pemecahan masalah yang menuntut tingkat kecermatan yang tinggi.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7d0gCxI38n9-yWR7cL3tbg.png" /><figcaption>Pemaparan materi praktis <em>cybersecurity</em> oleh Kak Dean Verrell Matalino Ariefin (Senior IT Network Specialist BCA). (Sumber: Dokumentasi)</figcaption></figure><p>Jika sesi tantangan teknis memberikan gambaran mengenai cara mencari celah keamanan, sesi selanjutnya, yaitu diskusi dan tanya jawab yang dipandu oleh Ibu Ilma Alpha Mannix, S.Kom., M.Kom. selaku dosen dan peneliti dari Fasilkom UI, menunjukkan implementasi wawasan tersebut di lapangan operasional perbankan. Dalam sesi ini, Kak Verrell menyoroti perbedaan yang cukup signifikan antara eksperimen keamanan di skala laboratorium dengan pengelolaan ribuan perangkat jaringan di sektor perbankan nasional. Di lingkungan ini, keamanan tidak dapat dipisahkan dari perancangan arsitektur jaringan yang matang. Beliau menekankan pentingnya penguasaan dokumen perancangan secara menyeluruh, yang mencakup HLD (<em>High Level Design</em>) untuk memetakan arsitektur dan topologi dari sistem makro, sampai LLD (<em>Low Level Design</em>) yang menjelaskan rincian konfigurasi teknis, <em>port</em>, dan aturan aliran data secara detail.</p><p>Melengkapi wawasan mengenai arsitektur tersebut, Kak Verrell juga memberikan <em>roadmap</em> bagi para mahasiswa yang ingin membangun kompetensi dengan standar industri. Beliau menunjukkan pentingnya sertifikasi keamanan jaringan terstandar, seperti sertifikasi Fortinet NSE (<em>Network Security Expert</em>) untuk memahami konsep <em>perimeter defense</em>, pengelolaan <em>firewall</em>, sampai keamanan jaringan terintegrasi yang profesional. Pemahaman mendalam atas alur komunikasi data dan konfigurasi jaringan adalah hal yang wajib dikuasai dan menjadi pembeda antara seorang praktisi profesional dengan orang yang hanya menjalankan alat tanpa mengerti logika di baliknya.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fIQWTRXSmFUTjfM4esSOog.png" /><figcaption>Sesi Talkshow dan Q&amp;A interaktif bersama Kak Dean Verrell Matalino Ariefin yang dipandu oleh moderator Ibu Ilma Alpha Mannix, S.Kom., M.Kom. (Sumber: Dokumentasi)</figcaption></figure><p>Seluruh ilmu dan wawasan yang dibedah dalam Seminar XCelerate Batch 2 menjadi salah satu langkah penting dalam ekosistem pembekalan talenta COMPFEST 18. Melalui rangkaian XCelerate, ditambah dengan sesi <em>workshop hands</em>-<em>on</em> bersama praktisi konsultan keamanan siber, COMPFEST 18 terus berupaya untuk membekali generasi muda dengan teori yang fundamental serta keahlian praktik. Pembekalan ini juga melengkapi pilar kompetisi CTF tingkat nasional serta program akademi yang dipersiapkan menuju puncak <em>Main Event</em> mendatang.</p><p>Bagi talenta muda yang ingin mempersiapkan diri dalam mengambil peran sebagai garda terdepan pelindung ekosistem digital Indonesia, pintu eksplorasi telah terbuka lebar. Pantau terus informasi jadwal dan rangkaian program berikutnya lewat situs web resmi dan media sosial COMPFEST agar tidak tertinggal dalam mengambil peran dalam transformasi teknologi saat ini. Mari persiapkan diri, asah potensi, dan <em>enclose the gap!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=31b8378099a6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/mendobrak-stereotip-peretasan-menemukan-esensi-proteksi-sistem-di-xcelerate-batch-2-compfest-18-31b8378099a6">Mendobrak Stereotip Peretasan: Menemukan Esensi Proteksi Sistem di XCelerate Batch 2 COMPFEST 18</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Lebih dari Sekadar Chatbot : AI dan Pembelajaran di Masa Depan]]></title>
            <link>https://medium.com/compfest/lebih-dari-sekadar-chatbot-ai-dan-pembelajaran-di-masa-depan-327530f4a1a1?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/327530f4a1a1</guid>
            <category><![CDATA[learning]]></category>
            <category><![CDATA[artificial-learning]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[aikou]]></dc:creator>
            <pubDate>Mon, 14 Sep 2026 13:50:33 GMT</pubDate>
            <atom:updated>2026-09-14T13:50:32.483Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3gsoRmHVaKPJ4Pfsjn_lxQ.jpeg" /></figure><p><a href="https://medium.com/@aikozhw/beyond-the-chatbots-ai-and-the-future-of-learning-3b414916ce84">Baca dalam Bahasa Inggris di sini!</a></p><p>Belajar menggunakan <em>Artificial Intelligence </em>selalu jadi topik yang sering dibicarakan oleh para siswa. Hal ini menunjukkan bahwa teknologi ini sudah sangat dikenal dalam kehidupan mereka. Namun, peran AI dalam dunia pendidikan terus berkembang dan tidak hanya sebatas menjadi <em>chatbot</em>. Salah satu contohnya adalah peningkatan fitur pembelajaran berbasis AI yang dikeluarkan oleh Google, termasuk Google Search, Gemini, dan Google Lens. Google mengatakan bahwa <em>update</em> ini menunjukkan bagaimana AI berkembang dari sekadar<em> chatbot</em> yang menjelaskan materi menjadi alat yang bisa membuat bahan belajar dan membantu siswa secara aktif selama proses <em>problem solving</em>.</p><p>Perkembangan ini menandai perubahan peran AI dalam pendidikan. Dulu AI hanya berfungsi sebagai alat yang menampilkan jawaban. Sekarang, AI mulai mendukung berbagai tahap dalam proses belajar. AI bisa membantu siswa mengatur informasi, membuat materi belajar, menguji pemahaman, memvisualisasikan konsep, dan memberikan petunjuk saat menyelesaikan soal.</p><p>AI juga bisa membuat proses belajar lebih menarik. Metode belajar tradisional biasanya mengandalkan buku teks dan mendengarkan penjelasan. Meski metode ini masih penting, AI bisa menawarkan cara lain agar siswa tetap fokus dan tidak merasa bosan. Contohnya adalah fitur Gemini. Ketika <em>user</em> meminta AI untuk menunjukkan bagaimana DNA bekerja dalam bentuk tiga dimensi, mereka bisa mengamati struktur DNA dengan cara memutar, memperbesar, atau memperkecil tampilannya. AI juga bisa memberi bantuan visual seperti tabel atau grid untuk menjelaskan konsep tertentu. Visualisasi ini bisa membuat materi yang rumit lebih mudah dipahami.</p><p>Pengembangan lainnya adalah peran AI dalam <em>problem solving</em>. AI tidak selalu harus langsung memberikan jawaban. Justru, AI bisa membantu menjelaskan konsep yang relevan, menunjukkan kemungkinan kesalahan, dan memberi petunjuk selama proses mencari solusi. Pendekatan ini bisa memberi manfaat lebih besar karena fokusnya bukan hanya pada jawaban akhir, tetapi juga pada proses mencari jawaban itu sendiri.</p><p>Namun, semakin banyak penggunaan AI dalam pendidikan juga membawa tantangan. Salah satunya adalah risiko ketergantungan berlebihan terhadap AI. Jika AI hanya digunakan sebagai jalan pintas, siswa bisa kehilangan kesempatan untuk melatih kemampuan berpikir kritis dan pemecahan masalah mereka sendiri.</p><p>Selain itu, akurasi informasi yang dihasilkan AI juga menjadi perhatian. AI bisa memberikan jawaban yang terdengar benar, tetapi sebenarnya tidak. Karena itu, siswa tidak boleh menerima informasi dari AI secara langsung. Kemampuan untuk memverifikasi informasi dari sumber terpercaya dan berpikir kritis tetap penting.</p><p>Pengembangan yang dilakukan Google adalah contoh dari perubahan yang sedang berlangsung dalam penggunaan AI untuk pendidikan. Secara luas, AI memiliki potensi untuk membantu siswa di berbagai tahap pembelajaran, mulai dari mencari informasi, memahami konsep, hingga berlatih dan menyelesaikan masalah.</p><p>Pada akhirnya, tujuan penggunaan AI dalam pendidikan seharusnya bukan untuk menggantikan proses belajar siswa. AI seharusnya menjadi alat yang membantu siswa memahami informasi secara lebih mendalam dan menjadi bagian aktif dalam proses pembelajaran. Penggunaan AI yang efektif bukanlah saat teknologi menggantikan kemampuan berpikir manusia, tetapi saat teknologi membantu siswa mengembangkan kemampuan tersebut.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=327530f4a1a1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/lebih-dari-sekadar-chatbot-ai-dan-pembelajaran-di-masa-depan-327530f4a1a1">Lebih dari Sekadar Chatbot : AI dan Pembelajaran di Masa Depan</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Beyond the Chatbots: AI and the Future of Learning]]></title>
            <link>https://medium.com/compfest/beyond-the-chatbots-ai-and-the-future-of-learning-3b414916ce84?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/3b414916ce84</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[learning]]></category>
            <dc:creator><![CDATA[aikou]]></dc:creator>
            <pubDate>Mon, 14 Sep 2026 13:50:21 GMT</pubDate>
            <atom:updated>2026-09-14T13:52:22.148Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3gsoRmHVaKPJ4Pfsjn_lxQ.jpeg" /></figure><p><a href="https://medium.com/@aikozhw/327530f4a1a1">Read in Indonesian here!</a></p><p>Learning with Artificial Intelligence has always been a topic among students which proves that it is a familiar part of students lives. However the role of AI in education is gradually expanding beyond conversations. One example of this development is Google’s recent expansion of AI- learning features across Google Search, Gemini and Google Lens. Google emphasized that these updates demonstrate how AI is evolving beyond a chatbot that simply explains content into a tool that can create study materials and actively support students during the problem-solving process.</p><p>This development shows a shift in the role of AI in education. Rather than functioning only as a tool that provides answers AI is beginning to support different stages of the learning process. It can help students organize information, create learning materials, test their understanding, visualize concepts and receive guidance while solving problems.</p><p>AI also has the potential to make learning more interactive. Traditional learning often involves reading textbooks, listening to lectures and completing assignments. While these methods are still important, AI can introduce ways for students to keep up with the materials without getting bored. Take a glance at what Gemini’s expanding. When a user asks to “show how DNA works in 3D” they can rotate, zoom in, and zoom out on the generated DNA structure to explore it directly. Responses also include question- visual aids such as tables and grids. Visual aids like that can help explain the underlying principles.</p><p>Another important development is AI’s role in the problem-solving process. AI does not necessarily have to provide students with the answer that way. This approach can be more beneficial for learning because it focuses not on the answer but also on the process of reaching it.</p><p>On the other hand there are always challenges from the growing role of AI in education. One of the concerns is overreliance on AI. If AI is used as a shortcut to complete academic tasks students may spend time developing their own critical thinking and problem-solving skills.</p><p>Another concern is the accuracy of AI-generated information. AI might provide inaccurate information that sounds legit. Hence students should not accept AI-generated content without questioning it. The need to verify information using sources and develop the ability to think critically are always needed for everyone.</p><p>Google’s recent development provides one example of this shift. The impact extends beyond a single company or platform. AI has the potential to support students throughout stages of learning from discovering information and understanding concepts to practicing, reviewing and solving problems.</p><p>However the purpose of AI in education should not be to do the learning for students. Instead it should help students understand information deeply and become active participants in their own learning process. The effective use of AI is not when it replaces thinking but when it supports students in developing it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3b414916ce84" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/beyond-the-chatbots-ai-and-the-future-of-learning-3b414916ce84">Beyond the Chatbots: AI and the Future of Learning</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Xcelerate Batch 1, From Data to Decisions: A Hands-On Introduction to Practical Machine Learning]]></title>
            <link>https://medium.com/compfest/xcelerate-batch-1-from-data-to-decisions-a-hands-on-introduction-to-practical-machine-learning-9faa355454b5?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/9faa355454b5</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[computer-vision]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[compfest]]></category>
            <category><![CDATA[object-detection]]></category>
            <dc:creator><![CDATA[Gde Maharta Putra]]></dc:creator>
            <pubDate>Wed, 19 Aug 2026 03:21:56 GMT</pubDate>
            <atom:updated>2026-08-19T03:21:55.267Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wv6kkEtlCnzA0R-jwVWYcw.png" /></figure><p><a href="https://medium.com/@gde.maharta.putra/f13c5b73044e">Baca Versi Bahasa Indonesia di Sini</a></p><p>Every morning, millions of Indonesians unlock their phones simply by looking at the screen. By midday, a parking barrier at the shopping mall lifts after a camera reads their licence plate. Come evening, their gallery app has neatly sorted holiday photos by the faces of the people closest to them. All of it feels perfectly ordinary, until one question surfaces: how exactly does a machine “see”?</p><p>That is precisely the question dissected in Xcelerate Batch 1, a seminar in the COMPFEST 18 series held online on Sunday, 2 August 2026. Carrying the theme “From Data to Decisions: A Hands-On Introduction to Practical Machine Learning”, the session was designed to carry participants across the gap that most often stops beginners in their tracks, namely the distance between reading about machine learning theory and actually running it themselves.</p><p>Xcelerate exists as a practical skills programme under the wider COMPFEST 18 banner, “Driving Infinite Growth to Enclose the Trench of Digital Skill Gap”. Where the Grand Launching session addressed the urgency of the digital skill gap at the level of ideas, Xcelerate Batch 1 answered it in the most concrete way possible, that is by inviting participants to open their laptops, write code, and train a model of their own making.</p><p><strong>Opening</strong></p><p>The event was opened by the COMPFEST 18 team, who welcomed participants from various regions as they said hello through the chat column. After running through the technical housekeeping, the committee introduced the keynote speaker, Viona Zatil Aqmar Kaleb, a Software Engineer with eight years of experience building production-scale applications.</p><p>Kak Viona’s track record spans roles as a fullstack and frontend engineer, mentoring junior engineers, and time spent within the Y Combinator startup ecosystem. She currently works at PT Bank Mandiri (Persero) Tbk. It was this industry-rooted background that kept the session grounded, because every concept was consistently brought back to one simple question: “what is this actually used for in the real world?”</p><p>The session was then divided into three parts, namely the material presentation, hands-on practice, and a question and answer segment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_5DmIOSmpyO9ch3Z-q5ogw.png" /><figcaption>“Workshop Timeline” slide and speaker introduction. (Source: Documentation)</figcaption></figure><p><strong>Main Session: Untangling the Terms We Keep Mixing Up</strong></p><p>Kak Viona opened the material by settling a classic confusion shared by almost every beginner, that is the difference between Artificial Intelligence, Machine Learning, and Deep Learning. The three were pictured as a single layered family. AI is the large umbrella, namely the ambition of making machines do things that would normally require human intelligence. Machine Learning is one of the routes towards that ambition, that is by letting machines learn from data on their own. Deep Learning, in turn, is the branch of Machine Learning that relies on multi-layered neural networks.</p><p>The turning point in the audience’s understanding arrived on the following slide, when Kak Viona explained that machine learning fundamentally inverts the way conventional programming works. In classical programming, the programmer supplies data and rules in order to produce answers. A spam filter, for instance, is built by writing the rule “if the word free appears, flag it as spam”. The problem is that spammers simply write “fr33” instead, and the rule is immediately blunted.</p><p>In machine learning, by contrast, we supply data and answers so that the machine itself discovers the rules. Rather than writing thousands of rules, we hand over a hundred thousand emails already labelled as spam or not, and let the algorithm find the pattern by itself.</p><p>“You do not write the rules. The algorithm discovers them on its own.”</p><p>Kak Viona also corrected the expectations so often shaped by science fiction films. The AI available today is applied AI, namely systems that are extraordinarily strong at one narrow task while being blind to everything else. A cat-detection model can only detect cats. It cannot play chess, let alone hold a conversation. Generalized AI, capable of understanding almost any human intellectual task, does not yet exist.</p><p>So why has AI only exploded now, when the theory is already decades old? The answer lies in three ingredients that finally became available at the same time, namely data growing exponentially thanks to the internet, hardware in the form of GPUs that turned out to be remarkably well suited to parallel computation, and frameworks and cloud platforms such as Google Colab, which offer free GPUs so that expensive equipment is no longer a prerequisite.</p><p>The material moved on to the three categories of machine learning, namely supervised learning, which learns from practice questions together with their answers, unsupervised learning, which searches for patterns without any answers, and reinforcement learning, which learns through trial and error much like playing a game. At this point, Kak Viona shared one piece of practical advice worth noting by anyone just starting out.</p><p>“Start from your question, not from the algorithm.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-1Masg0k2AGCttZPPF3gbw.png" /><figcaption>“ML flips the way programming works” slide (Source: Documentation)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DXifDdfh_0KvlrQ5g44Cog.png" /><figcaption>“Three categories, three ways of learning” slide (Source: Documentation)</figcaption></figure><p><strong>When Machines Learn to See</strong></p><p>Entering the computer vision segment, participants were led to recognise one fundamental fact, namely that to a computer, an image is nothing more than a matrix of numbers. A grayscale image is represented by a single number per pixel, ranging from 0 for black to 255 for white, whereas an RGB image requires three numbers per pixel, producing roughly 16.7 million possible colours. It is from that stack of numbers that a model learns to recognise edges, textures, and eventually shapes.</p><p>Kak Viona then mapped out the three principal computer vision tasks that already surround us every day. Classification powers Face Unlock, Instagram filters, and photo sorting in Google Photos. Detection underpins Tesla Autopilot, licence plate reading in parking areas, and people tagging on Facebook. Segmentation is used for medical diagnosis, background replacement on Zoom, and portrait mode on smartphones.</p><p>To explain how YOLO works, the object detection algorithm at the heart of the practical session, Kak Viona reached for an easily digestible analogy, namely an experienced security guard. At a single glance, that guard knows there are five visitors, knows they are on the second and third floors, and knows that the one standing in the corner warrants a closer look. Those three capabilities correspond to counting objects, locating them, and classifying their context.</p><p>The portion that landed closest to home came when Kak Viona presented computer vision applications in Indonesia, ranging from national ID verification with liveness detection when opening a digital bank account, licence plate reading at toll gates and parking areas, plant disease detection from photographs of leaves, and lung disease screening from chest X-rays, through to image-based product search in e-commerce.</p><p>One concept was singled out as the key takeaway participants should carry home, namely transfer learning. Training a model from scratch demands millions of images and expensive GPUs, whereas transfer learning allows us to take a model already trained on a vast dataset such as ImageNet and then adapt it to our own needs.</p><p>“It is like hiring someone with ten years of experience rather than a fresh graduate. All you need to do is give them a short onboarding on the company context.”</p><p>Rather than dwelling only on the bright side, Kak Viona also ran a brief demonstration titled “Making Data Bias Visible in 5 Minutes” using Teachable Machine. A model was trained on organic waste samples all photographed on a white table, then tested with similar objects placed on a wooden table. Its accuracy collapsed immediately. The conclusion cut deep, namely that the model had never actually learned to recognise waste at all. It had learned to recognise the background. Participants were left with a clear realisation that data bias is not a theoretical concern but a mistake that can occur within minutes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jExsy_7pZxeK_HEBxwcHdQ.png" /><figcaption>“Image = matrix of numbers,” CV applications in Indonesia, and a demonstration of data bias slide. (Source: Documentation)</figcaption></figure><p><strong>Ice Breaking Session</strong></p><p>Keeping participants energised across two hours is no small feat. The committee prepared a word-guessing game on the Wayground platform, challenging participants to guess the names of cities and countries. Everyone joined in and the answer board filled with guesses almost instantly. Some questions were answered correctly by the majority, while others successfully threw participants off and drew laughter in the chat column.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*N3NL4_7VYkRA5sX3L1LoZQ.png" /><figcaption>Ice-breaking session slide. (Source: Documentation)</figcaption></figure><p><strong>Hands-on Session: Building a Waste Type Detector</strong></p><p>This is the part that sets Xcelerate apart from the average seminar. Participants were guided to open a prepared Google Colab notebook and then build their own waste type detection system based on YOLO.</p><p>The process unfolded step by step. A YOLO model already trained on the 80 classes of the COCO dataset was loaded, and its outputs were then mapped into three waste categories, namely organic, inorganic, and hazardous waste, known in Indonesia as B3. Of those 80 classes, 58 objects were successfully mapped. Items such as bottles, backpacks, umbrellas, and skateboards fell into the recyclable inorganic category, while phones, laptops, and computer mice were grouped as hazardous waste because they contain toxic components and batteries.</p><p>Participants went a step further by running real-time detection through their webcams, complete with a continuous detection mode and an audio cue serving as an alert. Within a single short session, a concept that had until then existed only as boxes on a slide turned into bounding boxes moving in step with objects in front of the participants’ own cameras.</p><p>As a complement, Kak Viona also demonstrated Teachable Machine by training a cat and dog classifier on hundreds of image samples, showing that building a model does not always demand advanced coding skills.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZiTOIoljtv5gKMlF8NXZmQ.png" /><figcaption>“Waste Detector” Colab notebook and real-time detection results. (Source: Documentation)</figcaption></figure><p><strong>Question and Answer Session</strong></p><p>Participant enthusiasm was plain to see in the question and answer session held through Slido, with more than a dozen questions arriving in a steady stream. Among them was whether supervised, unsupervised, and reinforcement learning could be combined within a single model. Another raised a genuinely difficult technical challenge, namely how a system distinguishes a clear PET plastic bottle from an ordinary plastic bag whose shape has already been crushed or folded.</p><p>Other questions touched on implementation and ethics. One participant asked about the procedure for deploying a model from the training stage into a production environment, given that the Colab exercise had only scratched the surface. Participants also raised the possibility of computer vision technology being misused by bad actors, along with how such misuse might be prevented. A fair number requested recommendations for open learning resources to pursue the field further.</p><p>Closing the session, Kak Viona shared a list of learning resources she considers worth opening, from Kaggle Learn and Fast.ai for the free and practical route, to the Ultralytics documentation as the official YOLO reference, through to CS231n Stanford, Roboflow, and Kaggle Competitions for those wanting to go deeper. Her message was brief but important, namely to pick one and stick with it rather than attempting everything at once.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wZB_3NrJoOQqd2O2pKajZQ.png" /><figcaption>Q&amp;A session slide and learning resource slide. (Source: Documentation)</figcaption></figure><p><strong>Closing</strong></p><p>The event closed with words of thanks to all participants, the speaker, the sponsor, and every media partner who supported the smooth running of the activity. The committee reminded participants to complete the feedback form as a prerequisite for receiving their e-certificate, and shared that the presentation material and event recording would be distributed by email. The session ended with a group photo on Zoom.</p><p>Xcelerate Batch 1 proved that the distance between “curious” and “have actually built one” is narrower than it appears. In the space of two hours, participants moved from confusing AI, Machine Learning, and Deep Learning to a model genuinely running on their own screens. The COMPFEST 18 series continues with a range of academies, competitions, and the Main Event. Keep an eye on the official COMPFEST website and social media so you do not miss out, and let us enclose the gap together!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9faa355454b5" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/xcelerate-batch-1-from-data-to-decisions-a-hands-on-introduction-to-practical-machine-learning-9faa355454b5">Xcelerate Batch 1, From Data to Decisions: A Hands-On Introduction to Practical Machine Learning</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Xcelerate Batch 1, Dari Data ke Keputusan: Pengantar Praktis Untuk Machine Learning]]></title>
            <link>https://medium.com/compfest/xcelerate-batch-1-from-data-to-decisions-a-hands-on-introduction-to-practical-machine-learning-f13c5b73044e?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/f13c5b73044e</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[compfest]]></category>
            <category><![CDATA[computer-vision]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[object-detection]]></category>
            <dc:creator><![CDATA[Gde Maharta Putra]]></dc:creator>
            <pubDate>Wed, 19 Aug 2026 03:21:48 GMT</pubDate>
            <atom:updated>2026-08-19T03:36:29.974Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wv6kkEtlCnzA0R-jwVWYcw.png" /></figure><p><a href="https://medium.com/@gde.maharta.putra/9faa355454b5">Read English Version Here</a></p><p>Setiap pagi, jutaan orang Indonesia membuka kunci ponsel hanya dengan menatap layar. Siang harinya, palang parkir di pusat perbelanjaan terangkat setelah kamera membaca pelat nomor kendaraan mereka. Menjelang malam, aplikasi galeri dengan rapi mengelompokkan foto liburan berdasarkan wajah orang-orang terdekat. Semua terasa biasa saja, sampai muncul satu pertanyaan: sebenarnya, bagaimana mesin bisa “melihat”?</p><p>Pertanyaan itulah yang dibedah tuntas dalam Xcelerate Batch 1, seminar dalam<em> </em>rangkaian acara COMPFEST 18 yang digelar secara daring pada Minggu, 2 Agustus 2026. Mengusung tema “From Data to Decisions: A Hands-On Introduction to Practical Machine Learning”, sesi ini dirancang untuk membawa peserta melewati jurang yang paling sering menghentikan pemula, yakni jarak antara membaca teori machine learning dan benar-benar menjalankannya sendiri.</p><p>Xcelerate hadir sebagai program pembekalan keterampilan praktis di bawah payung besar COMPFEST 18, “Driving Infinite Growth to Enclose the Trench of Digital Skill Gap”. Jika sesi Grand Launching berbicara tentang urgensi digital skill gap di tataran gagasan, Xcelerate Batch 1 menjawabnya dengan cara yang paling konkret, yaitu mengajak peserta membuka laptop, menulis kode, dan melatih model buatan mereka sendiri.</p><p><strong>Pembukaan</strong></p><p>Acara dibuka oleh tim COMPFEST 18 yang menyambut peserta dari berbagai daerah yang menyapa lewat kolom chat. Setelah menyampaikan rangkaian teknis, panitia memperkenalkan pembicara utama, yaitu Viona Zatil Aqmar Kaleb, seorang Software Engineer dengan pengalaman delapan tahun membangun aplikasi berskala produksi.</p><p>Rekam jejak Kak Viona mencakup peran sebagai <em>fullstack </em>dan <em>frontend engineer</em>, mentor bagi engineer pemula, serta pengalaman di ekosistem startup Y Combinator. Ia kini berkarier di PT Bank Mandiri (Persero) Tbk. Latar belakang yang berakar pada dunia industri inilah yang membuat sesi berjalan sangat konkret, karena setiap konsep selalu dipulangkan ke satu pertanyaan sederhana, yakni “ini dipakai untuk apa di dunia nyata?”.</p><p>Sesi kemudian dibagi menjadi tiga bagian, yaitu pemaparan materi, praktik langsung, dan tanya jawab.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_5DmIOSmpyO9ch3Z-q5ogw.png" /><figcaption><em>Slide </em>“Timeline Workshop” dan perkenalan pembicara. (Sumber: Dokumentasi)</figcaption></figure><p><strong>Sesi Utama: Membongkar Istilah yang Selama Ini Tertukar</strong></p><p>Kak Viona membuka materi dengan menyelesaikan kebingungan klasik yang dialami hampir semua pemula, yaitu perbedaan antara <em>Artificial Intelligence</em>, <em>Machine Learning</em>, dan <em>Deep Learning</em>. Ketiganya digambarkan sebagai satu keluarga berlapis. AI adalah payung besarnya, yakni cita-cita membuat mesin melakukan hal yang biasanya membutuhkan kecerdasan manusia. <em>Machine Learning</em> merupakan salah satu jalan menuju cita-cita itu, yaitu dengan membiarkan mesin belajar sendiri dari data. Adapun <em>Deep Learning</em> adalah bagian dari <em>Machine Learning </em>yang memanfaatkan <em>neural network </em>berlapis-lapis.</p><p>Titik balik pemahaman peserta datang pada <em>slide </em>berikutnya, ketika Kak Viona menjelaskan bahwa <em>machine learning</em> sesungguhnya membalik cara kerja pemrograman konvensional. Pada pemrograman klasik, <em>programmer </em>memberi data dan aturan untuk menghasilkan jawaban. Contohnya, penyaring spam dibuat dengan menulis aturan “kalau ada kata gratis, tandai sebagai spam”. Masalahnya, para pengirim spam tinggal menulis “gr4tis” dan aturan itu langsung tumpul.</p><p>Sebaliknya, pada <em>machine learning</em>, kita justru memberi data dan jawaban agar mesin sendiri yang menemukan aturannya. Alih-alih menulis ribuan aturan, kita cukup menyodorkan seratus ribu surel yang sudah ditandai spam atau bukan, lalu membiarkan algoritma mencari polanya sendiri.</p><p>“Aturannya bukan kamu yang tulis, tetapi ditemukan sendiri oleh algoritma.”</p><p>Kak Viona juga meluruskan ekspektasi yang kerap dibentuk oleh film-film fiksi ilmiah. AI yang tersedia hari ini adalah <em>applied </em>AI, yakni sistem yang sangat hebat pada satu tugas sempit tetapi buta pada hal lain. Model pendeteksi kucing hanya bisa mendeteksi kucing, tidak bisa bermain catur, apalagi mengobrol. Sementara itu, <em>generalized </em>AI yang mampu memahami hampir semua tugas intelektual manusia masih belum ada.</p><p>Lalu, mengapa AI baru meledak sekarang padahal teorinya sudah berumur puluhan tahun? Jawabannya terletak pada tiga bahan yang akhirnya tersedia bersamaan, yaitu data yang tumbuh eksponensial berkat internet, hardware berupa GPU yang ternyata sangat cocok untuk komputasi paralel, serta framework dan cloud seperti Google Colab yang menyediakan GPU gratis sehingga tidak perlu lagi membeli perangkat mahal.</p><p>Materi berlanjut ke tiga kategori cara belajar mesin, yakni <em>supervised learning</em> yang belajar dari contoh soal beserta jawabannya, <em>unsupervised learning</em> yang mencari pola tanpa jawaban, dan <em>reinforcement learning</em> yang belajar lewat coba-coba layaknya bermain gim. Pada bagian ini, Kak Viona menyampaikan satu pesan praktis yang layak dicatat siapa pun yang baru memulai.</p><p>“Mulai dari pertanyaanmu, bukan dari algoritmanya.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-1Masg0k2AGCttZPPF3gbw.png" /><figcaption><em>Slide </em>“ML membalik cara kerja programming” (Sumber: Dokumentasi)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DXifDdfh_0KvlrQ5g44Cog.png" /><figcaption><em>Slide </em>“Tiga kategori, tiga cara belajar” (Sumber: Dokumentasi)</figcaption></figure><p><strong>Ketika Mesin Belajar Melihat</strong></p><p>Memasuki babak <em>computer vision</em>, peserta diajak menyadari satu fakta mendasar, yaitu bahwa bagi komputer, gambar hanyalah matriks angka. Citra <em>grayscale </em>diwakili satu angka per piksel dengan rentang 0 untuk hitam hingga 255 untuk putih, sedangkan citra RGB membutuhkan tiga angka per piksel sehingga menghasilkan sekitar 16,7 juta kemungkinan warna. Dari tumpukan angka itulah model belajar mengenali tepi, tekstur, dan akhirnya bentuk.</p><p>Kak Viona lalu memetakan tiga tugas utama <em>computer vision</em> yang sebenarnya sudah mengelilingi kita setiap hari. Klasifikasi bekerja pada <em>Face Unlock</em>, filter Instagram, dan penyortiran foto di Google Photos. Deteksi menopang Tesla Autopilot, pembacaan pelat nomor di area parkir, serta penandaan orang di Facebook. Segmentasi dipakai untuk diagnosis medis, penggantian latar belakang di Zoom, hingga mode potret pada ponsel.</p><p>Untuk menjelaskan cara kerja YOLO, algoritma deteksi objek yang menjadi inti sesi praktik, Kak Viona memakai analogi yang mudah dicerna, yaitu seorang pengawas keamanan berpengalaman. Dalam sekali pandang, pengawas tersebut langsung tahu ada lima pengunjung, tahu mereka berada di lantai dua dan tiga, sekaligus tahu bahwa yang berdiri di sudut ruangan patut dicurigai. Ketiga kemampuan itu setara dengan menghitung jumlah objek, menentukan lokasi objek, dan mengklasifikasikan konteksnya.</p><p>Bagian yang paling menyentuh konteks lokal muncul saat Kak Viona memaparkan penerapan <em>computer vision</em> di Indonesia, mulai dari verifikasi KTP dengan liveness detection pada pembukaan rekening digital, pembacaan pelat nomor di gerbang tol, deteksi penyakit tanaman melalui foto daun, penapisan penyakit paru dari citra rontgen dada, sampai fitur pencarian produk berbasis gambar di e-commerce.</p><p>Satu konsep ditegaskan sebagai bekal utama yang harus dibawa pulang peserta, yaitu transfer learning. Melatih model dari nol menuntut jutaan gambar dan GPU mahal, sedangkan transfer learning memungkinkan kita memakai model yang sudah dilatih pada dataset raksasa seperti ImageNet, lalu menyesuaikannya dengan kebutuhan sendiri.</p><p>“Seperti merekrut karyawan yang sudah berpengalaman sepuluh tahun, bukan <em>fresh graduate</em>. Kamu tinggal memberi onboarding singkat soal konteks perusahaan.”</p><p>Tidak berhenti pada sisi cerahnya, Kak Viona juga menggelar demonstrasi singkat bertajuk “Bikin Bias Data Terlihat dalam 5 Menit” menggunakan Teachable Machine. Sebuah model dilatih dengan seluruh contoh sampah organik yang difoto di atas meja putih, lalu diuji dengan objek serupa di atas meja kayu. Akurasinya langsung anjlok. Kesimpulannya terasa menohok, yakni model tersebut sebenarnya tidak pernah belajar mengenali sampah, melainkan belajar mengenali latar belakang. Peserta pun diajak menyadari bahwa bias data bukan isu teoretis, melainkan kesalahan yang bisa terjadi dalam hitungan menit.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jExsy_7pZxeK_HEBxwcHdQ.png" /><figcaption><em>Slide </em>“Gambar = matriks angka”, penerapan CV di Indonesia, dan demonstrasi bias data. (Sumber: Dokumentasi)</figcaption></figure><p><strong>Sesi <em>Ice Breaking</em></strong></p><p>Menjaga energi peserta selama dua jam bukan perkara mudah. Panitia menyiapkan permainan tebak kata melalui <em>platform </em>Wayground, dengan tantangan menebak nama kota dan negara. Seluruh peserta terlibat aktif dan papan jawaban langsung dipenuhi tebakan. Beberapa soal dijawab benar oleh mayoritas peserta, sementara soal lain berhasil mengecoh dan memancing gelak tawa di kolom chat.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*N3NL4_7VYkRA5sX3L1LoZQ.png" /><figcaption><em>Slide </em>sesi ice breaking. (Sumber: Dokumentasi)</figcaption></figure><p><strong>Sesi <em>Hands-on</em>: Membangun Pendeteksi Jenis Sampah</strong></p><p>Inilah bagian yang membedakan Xcelerate dari seminar pada umumnya. Peserta diarahkan membuka <em>notebook </em>Google Colab yang telah disiapkan, lalu membangun sendiri sebuah sistem pendeteksi jenis sampah berbasis YOLO.</p><p>Prosesnya berjalan bertahap. Model YOLO yang telah dilatih pada 80 kelas dataset COCO dimuat, kemudian keluarannya dipetakan ke dalam tiga kategori sampah, yaitu organik, anorganik, dan B3 alias bahan berbahaya dan beracun. Dari 80 kelas tersebut, 58 objek berhasil dipetakan. Barang seperti botol, ransel, payung, dan papan seluncur masuk kategori anorganik yang masih bisa didaur ulang, sedangkan ponsel, laptop, dan tetikus dikelompokkan sebagai B3 karena mengandung komponen beracun dan baterai.</p><p>Peserta melangkah lebih jauh dengan menjalankan deteksi waktu nyata melalui webcam, lengkap dengan mode deteksi berkelanjutan dan penanda suara sebagai peringatan. Dalam satu sesi singkat, konsep yang tadinya hanya berupa kotak-kotak pada <em>slide </em>berubah menjadi kotak pembatas yang bergerak mengikuti objek di depan kamera peserta sendiri.</p><p>Sebagai pelengkap, Kak Viona juga mendemonstrasikan Teachable Machine untuk melatih pengklasifikasi kucing dan anjing dengan ratusan sampel gambar, memperlihatkan bahwa membangun model tidak selalu menuntut penguasaan kode tingkat lanjut.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZiTOIoljtv5gKMlF8NXZmQ.png" /><figcaption>Notebook Colab “Pendeteksi Sampah” dan hasil deteksi real-time. (Sumber: Dokumentasi)</figcaption></figure><p><strong>Sesi Tanya Jawab</strong></p><p>Antusiasme peserta terlihat jelas pada sesi tanya jawab yang berlangsung melalui Slido, dengan belasan pertanyaan yang terus berdatangan. Pertanyaan-pertanyaan yang muncul seperti apakah metode <em>supervised</em>, <em>unsupervised</em>, dan <em>reinforcement learning</em> dapat digabungkan dalam satu model. Terdapat juga pertanyaan terkait tantangan teknis, yaitu bagaimana sistem membedakan botol plastik bening berbahan PET dari kantong plastik biasa yang bentuknya sudah hancur atau terlipat.</p><p>Pertanyaan lain menyentuh sisi penerapan dan etika. Ada peserta yang menanyakan prosedur <em>deployment model</em> dari tahap latihan menuju lingkungan produksi, mengingat praktik di Colab baru menyentuh permukaan. Peserta menyoroti kemungkinan penyalahgunaan teknologi <em>computer vision</em> oleh pihak tidak bertanggung jawab beserta cara mencegahnya. Tidak sedikit pula yang meminta rekomendasi sumber belajar terbuka untuk mendalami bidang ini.</p><p>Menutup sesi, Kak Viona membagikan daftar sumber belajar yang menurutnya layak dibuka, mulai dari Kaggle Learn dan Fast.ai untuk jalur gratis dan praktis, dokumentasi Ultralytics sebagai rujukan resmi YOLO, hingga CS231n Stanford, Roboflow, dan Kaggle Competitions bagi yang ingin melangkah lebih jauh. Pesannya singkat namun penting, yakni pilih satu dan tekuni, jangan mencoba semuanya sekaligus.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wZB_3NrJoOQqd2O2pKajZQ.png" /><figcaption><em>Slide </em>sesi tanya jawab. (Sumber: Dokumentasi)</figcaption></figure><p><strong>Penghujung Acara</strong></p><p>Acara ditutup dengan ucapan terima kasih kepada seluruh peserta, pembicara, pihak sponsor, serta seluruh media partner yang telah mendukung kelancaran kegiatan. Panitia mengingatkan peserta untuk mengisi formulir umpan balik sebagai prasyarat penerimaan e-sertifikat, sekaligus menyampaikan bahwa materi presentasi dan rekaman acara akan dibagikan melalui surel. Sesi ditutup dengan dokumentasi bersama di Zoom.</p><p>Xcelerate Batch 1 membuktikan bahwa jarak antara “penasaran” dan “pernah membuat” tidak selebar yang dibayangkan. Dalam dua jam, peserta bergerak dari kebingungan membedakan AI, <em>Machine Learning</em>, dan <em>Deep Learning </em>menuju sebuah model yang benar-benar berjalan di layar mereka sendiri. Rangkaian COMPFEST 18 masih akan berlanjut dengan berbagai akademi, kompetisi, dan Main Event. Pantau terus situs web resmi serta media sosial COMPFEST agar tidak tertinggal, dan mari bersama-sama <em>enclose the gap</em>!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f13c5b73044e" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/xcelerate-batch-1-from-data-to-decisions-a-hands-on-introduction-to-practical-machine-learning-f13c5b73044e">Xcelerate Batch 1, Dari Data ke Keputusan: Pengantar Praktis Untuk Machine Learning</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[XCelerate Foundations : Terperangkap Dalam FYP — Menjembatani Kesenjangan Antara Pengguna…]]></title>
            <link>https://medium.com/compfest/xcelerate-foundations-trapped-in-the-fyp-bridging-the-gap-between-everyday-users-and-big-data-356c91fd5d46?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/356c91fd5d46</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[social-media]]></category>
            <category><![CDATA[big-data]]></category>
            <dc:creator><![CDATA[justinevan]]></dc:creator>
            <pubDate>Wed, 15 Jul 2026 09:27:28 GMT</pubDate>
            <atom:updated>2026-08-19T03:34:49.115Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>XCelerate Foundations : Terperangkap Dalam FYP — Menjembatani Kesenjangan Antara Pengguna Sehari-hari dan Big Data</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/869/1*MRyIalY4GOSDV0Ci1tzTGg.jpeg" /></figure><p><a href="https://medium.com/@jhalimsaputra/xcelerate-foundations-trapped-in-the-fyp-bridging-the-gap-between-everyday-users-and-big-data-6d7a50ed4452">Baca Dalam Bahasa Inggris</a></p><p>Berawal dari niat beristirahat sejenak setelah beraktivitas, kita sering kali kebablasan <em>scrolling</em> media sosial hingga berjam-jam. Fenomena yang biasa disebut <em>doom scrolling</em> ini membuat kita lupa waktu dan akhirnya menjadi tidak produktif. Memang sulit untuk tidak tenggelam dalam dunia “FYP” yang terus menyajikan video-video hiburan. Namun, jika terlalu lama hanyut di dalamnya, kita akan terjebak tanpa sempat menyusun rencana pengembangan diri untuk masa depan. Alhasil, kesenjangan antara kompetensi diri kita saat ini dan tuntutan dunia kerja akan semakin melebar.</p><p>XCelarate Foundations merupakan program satu hari yang didesain khusus bagi individu yang tertarik memulai karier di bidang teknologi. Melalui seminar dan lokakarya (<em>workshop</em>) yang inspiratif, acara ini membuka pintu bagi siapa saja untuk memahami fundamental IT. Program ini hadir sebagai jembatan yang siap memperkecil kesenjangan antara kualitas diri kita saat ini dan realitas dunia kerja yang nyata.</p><h3>Pembukaan</h3><p>Acara dibuka dengan sesi interaktif oleh dua MC, yaitu Fatma dan Michelle. Mereka memperkenalkan tema seminar hari ini, yakni “Trapped in The FYP: Bridging the Gap Between Everyday Users and Big Data”. Selanjutnya, kedua MC turut memperkenalkan pembicara luar biasa yang hadir, yaitu Rashad Aziz (Data Scientist di Gojek) dan Muhammad Luthfi Reynaldi (Data Scientist di GoTo).</p><h3>Sesi Utama</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*u7xFzezmx7yI3IyZw9iZpw.jpeg" /><figcaption>Moderator — Gunta</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*G5LYuw-_5fpbUXQFmnlN3A.jpeg" /><figcaption>Pembicara 1 — Rashad Aziz</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JdKLCLvxSiUmtjoNlzRvpQ.jpeg" /><figcaption>Pembicara 2 — Muhammad Luthfi Reynaldi</figcaption></figure><p>Pada sesi utama, moderator melanjutkan diskusi dengan memberikan beberapa pertanyaan kepada Kak Rashad dan Kak Luthfi,</p><blockquote><em>“Apakah pengalaman yang dipelajari dari mata kuliah seperti Struktur Data dan Aljabar Linear terbawa ke penelitian LLM (Large Language Model)? “</em></blockquote><p>Mata kuliah Aljabar Linear dan Struktur Data sangat membantu membangun fondasi kuat untuk riset LLM. Menurut Kak Rashad, kesenjangan dengan industri dapat dijembatani oleh rasa ingin tahu yang tinggi (<em>stay curious</em>) demi mengimbangi pesatnya perkembangan teknologi. Menyambung hal itu, Kak Luthfi menambahkan bahwa <em>passion</em> dan rasa ingin tahu di bidang <em>Data Science</em> sangat krusial, di mana praktisi harus berani memecahkan masalah lewat eksperimen (<em>A/B testing</em>) tanpa takut gagal.</p><blockquote><em>“Bagaimana sistem rekomendasi, seperti di TikTok atau Amazon, seolah-olah bisa</em></blockquote><p>Sistem rekomendasi bekerja dengan melacak tindakan digital pengguna seperti klik, <em>scroll</em>, dan pesanan di <em>database backend</em>. Kak Luthfi menjelaskan bahwa algoritma akan mengidentifikasi pola perilaku dari jutaan pengguna ini untuk meningkatkan <em>engagement</em> dan transaksi bisnis. Menurut Kak Rashad, manusia pada dasarnya sangat mudah ditebak sehingga algoritma bisa terus beradaptasi dari umpan balik harian untuk memahami penggunanya dengan sempurna.</p><blockquote><em>“Kebiasaan atau faktor pengguna spesifik apa yang paling memengaruhi kategorisasi algoritma di perusahaan seperti GoTo/GoPay?”</em></blockquote><p>Kak Rashad mengungkapkan bahwa model AI sering kali bekerja sebagai <em>black box</em>, di mana pengembang sendiri tidak selalu tahu alasan pasti di balik suatu rekomendasi. Menyambung hal itu, Kak Luthfi menegaskan adanya tata kelola privasi data yang sangat ketat di industri sehingga data sensitif tidak bisa digunakan tanpa izin khusus. Alhasil, perusahaan lebih mengandalkan titik data umum untuk membentuk hipotesis luas dan mengambil keputusan bisnis melalui eksperimen.</p><blockquote><em>“Apa hambatan (bottleneck) terbesar saat beralih dari data mentah ke rekomendasi pengguna secara langsung?”</em></blockquote><p>Kak Luthfi menyoroti bahwa prapemrosesan data (<em>data pre-processing</em>) menjadi hambatan terbesar karena sebagian besar waktu habis hanya untuk membersihkan data mentah. Kak Rashad sepakat bahwa kebersihan data sangat krusial demi menghindari prinsip <em>“garbage in, garbage out”</em>. Selain itu, pergeseran perilaku pengguna seiring waktu (<em>feature drift</em>) menjadi tantangan besar lain yang mengharuskan tim memantau kapan model perlu dilatih ulang.</p><blockquote><em>“Apakah Filter Bubble, di mana pengguna diisolasi sehingga hanya melihat konten yang direkomendasikan secara spesifik, adalah fenomena nyata?”</em></blockquote><p>Fenomena <em>filter bubble</em> tercipta karena media sosial menyajikan konten yang sejalan dengan pandangan pengguna agar mereka terus melakukan <em>scrolling</em>. Kak Luthfi menjelaskan bahwa hal ini terjadi jika model hanya fokus mengejar klik <em>(exploitation</em>) tanpa menyisipkan konten baru (<em>exploration</em>). Guna mengatasinya, platform biasanya menyuntikkan konten acak dan Kak Rashad mengingatkan pengguna agar sadar diri serta aktif memberikan umpan balik.</p><blockquote><em>“Keterampilan apa yang paling layak dikejar saat ini bagi seseorang yang ingin memasuki dunia kerja data?”</em></blockquote><p>Kak Luthfi menekankan bahwa dunia data bukan sekadar soal keahlian teknis, melainkan kemampuan meyakinkan pemangku kepentingan lewat dampak bisnis yang nyata. Kak Rashad sangat sepakat karena hard skill akan terus berubah, sehingga kemampuan komunikasi kepada pihak non-teknis menjadi sangat vital. Akhirnya, keduanya menyimpulkan bahwa calon praktisi wajib mengikuti tren, mampu merumuskan hipotesis dengan baik, dan memiliki mental baja saat menghadapi kegagalan eksperimen.</p><h3>Sesi Game Interaktif</h3><p>Untuk menjaga semangat para peserta, kedua MC memimpin sesi permainan tebak kata yang seru melalui platform Quizizz.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*42_syvXLpJpkHoXYD02jmg.jpeg" /></figure><h3><strong>Sesi Tanya Jawab</strong></h3><p>Usai permainan, acara dilanjutkan dengan sesi tanya jawab. Para peserta menunjukkan antusiasme yang tinggi, terlihat dari banyaknya pertanyaan yang terus bermunculan melalui platform Slido.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/749/1*e8Ip1-o_-79eZYe3ImIvEQ.jpeg" /></figure><h3><strong>Penghujung Acara</strong></h3><p>Untuk menutup acara, MC menyampaikan rasa terima kasih kepada para peserta yang hadir, pihak sponsor, serta <em>media partner</em> yang telah mendukung kelancaran acara ini. MC juga mengingatkan bahwa masih banyak rangkaian acara menarik lainnya yang akan berlanjut di COMPFEST 18. Pada akhirnya, acara resmi ditutup dengan sesi dokumentasi bersama seluruh peserta di Zoom.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=356c91fd5d46" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/xcelerate-foundations-trapped-in-the-fyp-bridging-the-gap-between-everyday-users-and-big-data-356c91fd5d46">XCelerate Foundations : Terperangkap Dalam FYP — Menjembatani Kesenjangan Antara Pengguna…</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[XCelerate Foundations : Trapped in The FYP — Bridging the Gap Between Everyday Users and Big Data]]></title>
            <link>https://medium.com/compfest/xcelerate-foundations-trapped-in-the-fyp-bridging-the-gap-between-everyday-users-and-big-data-6d7a50ed4452?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/6d7a50ed4452</guid>
            <category><![CDATA[social-media]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[big-data]]></category>
            <dc:creator><![CDATA[justinevan]]></dc:creator>
            <pubDate>Wed, 15 Jul 2026 09:27:19 GMT</pubDate>
            <atom:updated>2026-07-15T09:27:17.633Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/869/1*MRyIalY4GOSDV0Ci1tzTGg.jpeg" /></figure><p><a href="https://medium.com/compfest/xcelerate-foundations-trapped-in-the-fyp-bridging-the-gap-between-everyday-users-and-big-data-356c91fd5d46">Read In Bahasa Indonesia</a></p><p>It often starts with a simple intention to take a short break after a busy day, but we frequently end up losing track of time, scrolling through social media for hours. This phenomenon, commonly known as doom scrolling, causes us to lose track of time and ultimately makes us unproductive. Admittedly, it is hard not to get drawn into the world of the “FYP” (For You Page) as it continuously serves up entertaining videos. However, if we drift away in it for too long, we will get trapped without ever making a self-development plan for the future. As a result, the gap between our current skills and the demands of the professional world will only widen.</p><p>XCelerate Foundations is a one-day program specifically designed for individuals interested in launching a career in the tech industry. Through inspiring seminars and workshops, this event opens doors for anyone wishing to understand IT fundamentals. The program serves as a bridge ready to close the gap between our current capabilities and the actual realities of the workforce.</p><h3><strong>Opening</strong></h3><p>The event kicked off with an interactive session hosted by two MCs, Fatma and Michelle. They introduced today’s seminar theme: “Trapped in The FYP: Bridging the Gap Between Everyday Users and Big Data.” Following that, the MCs introduced the remarkable speakers of the session: Rashad Aziz (Data Scientist at Gojek) and Muhammad Luthfi Reynaldi (Data Scientist at GoTo).</p><h3><strong>Main Session</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*u7xFzezmx7yI3IyZw9iZpw.jpeg" /><figcaption>Moderator — Gunta</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*G5LYuw-_5fpbUXQFmnlN3A.jpeg" /><figcaption>Speaker 1 — Rashad Aziz</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JdKLCLvxSiUmtjoNlzRvpQ.jpeg" /><figcaption>Speaker 2 — Muhammad Luthfi Reynaldi</figcaption></figure><p>During the main session, the moderator led the discussion by posing several questions to Kak Rashad and Kak Luthfi,</p><blockquote><em>“Does the experience gained from courses like Data Structures and Linear Algebra carry over into LLM (Large Language Model) research?”</em></blockquote><p>Linear Algebra and Data Structures courses are highly beneficial in building a solid foundation for LLM research. According to Kak Rashad, the gap between academia and the industry can be bridged by maintaining a high level of curiosity (stay curious) to keep pace with rapid technological advancements. Echoing this sentiment, Kak Luthfi added that passion and curiosity in Data Science are crucial, emphasizing that practitioners must be brave enough to solve problems through experimentation (A/B testing) without the fear of failure.</p><blockquote><em>“How do recommendation systems, like those on TikTok or Amazon, seem to read our minds and know exactly what to show us?”</em></blockquote><p>Recommendation systems work by tracking users’ digital footprints, such as clicks, scrolls, and orders, within a backend database. Kak Luthfi explained that algorithms identify behavioral patterns from millions of users to boost engagement and drive business transactions. According to Kak Rashad, humans are inherently highly predictable, allowing algorithms to continuously adapt based on daily feedback to understand their users perfectly.</p><blockquote><em>“What specific user habits or factors most heavily influence algorithmic categorization at companies like GoTo/GoPay?”</em></blockquote><p>Kak Rashad revealed that AI models often operate as a “black box,” where even developers do not always know the exact reasoning behind a specific recommendation. Building on that, Kak Luthfi emphasized that the industry enforces highly strict data privacy governance, meaning sensitive data cannot be utilized without special permission. As a result, companies rely more on generalized data points to formulate broad hypotheses and make business decisions through experimentation.</p><blockquote><em>“What is the biggest bottleneck when transitioning from raw data to real-time user recommendations?”</em></blockquote><p>Kak Luthfi highlighted that data pre-processing is the biggest bottleneck because a massive portion of the time is spent merely cleaning raw data. Kak Rashad agreed that data cleanliness is absolutely crucial to avoid the classic “garbage in, garbage out” pitfall. Additionally, the shift in user behavior over time (feature drift) poses another major challenge, requiring teams to carefully monitor when models need to be retrained.</p><blockquote><em>“Is the ‘Filter Bubble’, where users are isolated to only see specifically recommended content, a real phenomenon?”</em></blockquote><p>The filter bubble phenomenon occurs because social media platforms dish out content that aligns with a user’s existing views to keep them scrolling. Kak Luthfi explained that this happens when a model solely focuses on chasing clicks (exploitation) without introducing fresh, new content (exploration). To counter this, platforms usually inject random pieces of content into the feed, and Kak Rashad reminded users to stay self-aware and actively provide feedback.</p><blockquote><em>“What skills are most worth pursuing right now for someone looking to enter the data workforce?”</em></blockquote><p>Kak Luthfi stressed that the world of data is not just about technical expertise, but rather the ability to convince stakeholders through tangible business impact. Kak Rashad strongly agreed, noting that hard skills will continuously evolve, making communication with non-technical parties absolutely vital. Ultimately, both concluded that aspiring practitioners must keep up with trends, possess strong hypothesis formulation skills, and maintain a resilient mindset when facing failed experiments.</p><h3><strong>Interactive Game Session</strong></h3><p>To keep the energy high among participants, the two MCs led an exciting word-guessing game session through the Quizizz platform.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*42_syvXLpJpkHoXYD02jmg.jpeg" /></figure><h3><strong>Q&amp;A Session</strong></h3><p>Following the game, the event proceeded to a Q&amp;A session. The participants demonstrated high enthusiasm, evident from the flood of questions popping up on the Slido platform.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/749/1*e8Ip1-o_-79eZYe3ImIvEQ.jpeg" /></figure><h3><strong>Closing of the Event</strong></h3><p>To bring the event to a close, the MCs expressed their gratitude to the attending participants, sponsors, and media partners who supported the smooth running of the event. The MCs also reminded everyone that many more exciting event lineups would continue throughout COMPFEST 18. Finally, the event officially concluded with a group photo/documentation session with all participants on Zoom.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6d7a50ed4452" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/xcelerate-foundations-trapped-in-the-fyp-bridging-the-gap-between-everyday-users-and-big-data-6d7a50ed4452">XCelerate Foundations : Trapped in The FYP — Bridging the Gap Between Everyday Users and Big Data</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Grand Launching of COMPFEST 18: Bridging the Digital Skill Gap and Building Adaptive Talents in…]]></title>
            <link>https://medium.com/compfest/grand-launching-of-compfest-18-bridging-the-digital-skill-gap-and-building-adaptive-talents-in-the-8a1fca80fde3?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/8a1fca80fde3</guid>
            <category><![CDATA[competition]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[indonesia]]></category>
            <dc:creator><![CDATA[Malvin]]></dc:creator>
            <pubDate>Sun, 17 May 2026 15:32:36 GMT</pubDate>
            <atom:updated>2026-05-18T05:05:33.080Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Grand Launching of COMPFEST 18: Bridging the Digital Skill Gap and Building Adaptive Talents in the AI Era</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HEpdRccp0AR-zH4UfWEtEg.jpeg" /></figure><p><a href="https://medium.com/compfest/grand-launching-compfest-18-menjembatani-digital-skill-gap-mencetak-talenta-adaptif-di-era-ai-8afe55bd1eae">Baca Versi Bahasa Indonesia di Sini</a></p><p>With AI and technology evolving so fast, we’re currently facing a real crisis: the digital skill gap. This gap can seriously slow down innovation and productivity across different sectors, creating a huge disconnect between the high demand for modern tech skills and the actual talents available.</p><p>Stepping up to tackle this growing global challenge, the computer science students of Universitas Indonesia (Fasilkom UI) are bringing back the biggest student-held IT event of the year: COMPFEST 18. Officially kicking off with the theme “Driving Infinite Growth to Enclose the Trench of Digital Skill Gap”, COMPFEST is here to make sure Indonesia’s younger generation isn’t just digitally literate, but also adaptable enough to lead the tech transformation and compete in the automation era.</p><p>In response to COMPFEST 18, the Dean of Fasilkom UI, Prof. Dr. Eng. Wisnu Jatmiko, S.T., M.Kom., expressed his full support. In his remarks, he emphasized the essence of COMPFEST, stating that it is not merely a routine work program, but rather an intellectual legacy. “Through the competitions, academies, and exploration spaces prepared in COMPFEST 18, we are striving to produce digital talents who are not only technically proficient but also adaptive and solution-oriented in answering the needs of the times,” he asserted. This vision was echoed by the Chairman of BEM Fasilkom UI 2026, Makarim Zufar Prambudyo, who highlighted the importance of equipping the younger generation to transform from mere spectators into drivers of solutions.</p><p>Complementing both statements, the Project Officer of COMPFEST 18, Khansa Dinda Arya Putri, also explained the significance of organizing COMPFEST, which is expected to close the digital skill gap. According to her, the annual event, which began in 2009, serves as a tangible manifestation of Fasilkom UI students’ contribution to advancing the technology ecosystem in Indonesia for the wider community.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ln181HzZd4grBLZBd_Tojg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9Fw-RiwKn_4GowyZFFdrUw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d7abZgPlfMyMiZKWRLhXaA.png" /><figcaption>The opening remarks session of the COMPFEST 18 Grand Launching by Prof. Dr. Eng. Wisnu Jatmiko, S.T., M.Kom (Dean of Fasilkom UI), Makarim Zufar Prambudyo (Chairman of BEM Fasilkom UI 2026), and Kansa Dinda Arya Putri (Project Officer of COMPFEST 18). (Source: Documentation)</figcaption></figure><p>Interestingly, the main thing you need to survive today’s tech boom isn’t just the ability to write code. This was discussed heavily during the Keynote Speech session, which totally broke the old stereotypes about IT folks. Mr. Kristiyanto from Google for Education Indonesia opened the session by talking about how our “unique selling points” as talents are shifting. Now that AI can easily handle analytical and problem-solving tasks, human “superpowers” are actually shifting toward communication, presentation, and persuasion.</p><p>Continuing the talk on AI disruption, tech and AI expert Mr. Risman Adnan, Ph.D., introduced the concept of AI Native Engineering. Instead of seeing AI as a job-stealing threat, he emphasized that AI is an absolute must-have tool to speed up innovation. “AI isn’t the threat, but people who don’t use AI will be very threatened by this technology,” he reminded everyone.</p><p>The Keynote Speech wrapped up with cybersecurity practitioner Mr. Alfons Tanujaya, who gave a witty reality check on why agility is everything. Drawing from his own success in IT despite having an Economics background, he compared today’s industry to natural selection: “It’s not the T-Rex that survived until today; it’s the cockroach because it adapts. The winner isn’t the strongest, but the one who can adjust.” He also added that with cyber threats getting smarter, IT talents need to step up as the front line to protect our digital ecosystem.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pgBwktII1_0pZ564JacYBw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MCkIr2u9QfHUZ1frw0bGaQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Hmkw0U5vJWQFJHW075dgVA.png" /><figcaption>The lineup for the Keynote Speech session: Mr. Kristiyanto (Adaption Lead at Google for Education Indonesia), Mr. Risman Adnan, Ph.D. (Technology &amp; AI Expert), and Mr. Alfons Tanujaya (IT &amp; Financial Security Specialist at vaksin.com). (Source: Documentation)</figcaption></figure><p>If the Keynote Speech felt a bit conceptual, the next session brought things right down to earth. The Talkshow, titled “Reality Check: Bridging the Gap Between What You Know and What the Industry Demands,” was guided by Ms. Kamila Alifia Imanuddin, S.Kom., M.T.I. as the moderator. It unpacked the gap between campus theories and actual workplace demands, featuring Mr. Muhammad Mudrik, a Software Engineer at Flip, and Mr. Abdul Hasan Zawawi, an AI &amp; Machine Learning Engineer at Prudential.</p><p>Mr. Hasan pointed out a huge difference between campus and corporate life. In college, you’re graded on getting the right answers. But at work, your solutions have to be fast and actually make a real impact on the business. He reminded students not to just obsess over their GPA, but to build up their portfolios with real projects and sharpen the soft skills that machines just can’t copy.</p><p>Meanwhile, Mr. Mudrik talked about the dilemma we all face with how easy tech makes things today. Sure, AI can help you write code or finish tasks, but understanding the fundamentals, like logic and data structures, is what separates a truly reliable engineer from someone who just takes shortcuts. “Make AI your assistant to learn the concepts, not a shortcut to finish your homework,” he stressed. He added that, at the end of the day, what really matters when executing projects in the industry is how well you communicate and collaborate with your team, regardless of how great your hard skills are.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3nEX_ZLtxius4n04TdzJLw.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UdOUAHBOd6mqyLKSi3zXGg.jpeg" /><figcaption>The “Reality Check” Talkshow session with Ms. Kamila Alifia Imanuddin, S.Kom., M.T.I., Mr. Muhammad Mudrik (Software Engineer at Flip), and Mr. Abdul Hasan Zawawi (AI &amp; Machine Learning Engineer at Prudential). (Source: Documentation)</figcaption></figure><p>All these awesome insights from the Keynote Speech and Talkshow basically kick off the whole COMPFEST 18 journey. To meet real industry demands and give the younger generation the tools they need for the workplace, COMPFEST 18 has lined up some amazing flagship programs. Participants can deep-dive into technical and managerial skills through four main academies: Data Science Academy, Software Engineering Academy, User Experience Academy, and Product Management Academy.</p><p>Plus, there’s plenty of room to innovate through competitions like Competitive Programming, Capture the Flag, and Business IT Case. Everything wraps up at the Main Event, which features interactive experiences and a Job Fair to connect great talent with dream companies.</p><p>So, keep an eye on the official website and COMPFEST’s social media for registration info and schedules so you don’t miss out on taking part in today’s digital transformation! Let’s get ready, sharpen our skills, and don’t forget to enclose the gap!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8a1fca80fde3" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/grand-launching-of-compfest-18-bridging-the-digital-skill-gap-and-building-adaptive-talents-in-the-8a1fca80fde3">Grand Launching of COMPFEST 18: Bridging the Digital Skill Gap and Building Adaptive Talents in…</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Grand Launching COMPFEST 18: Menjembatani Digital Skill Gap, Mencetak Talenta Adaptif di Era AI]]></title>
            <link>https://medium.com/compfest/grand-launching-compfest-18-menjembatani-digital-skill-gap-mencetak-talenta-adaptif-di-era-ai-8afe55bd1eae?source=rss----8984eec5fa5d---4</link>
            <guid isPermaLink="false">https://medium.com/p/8afe55bd1eae</guid>
            <category><![CDATA[competition]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[indonesia]]></category>
            <dc:creator><![CDATA[Malvin]]></dc:creator>
            <pubDate>Sun, 17 May 2026 15:32:24 GMT</pubDate>
            <atom:updated>2026-05-18T05:06:17.721Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HEpdRccp0AR-zH4UfWEtEg.jpeg" /></figure><p><a href="https://medium.com/compfest/grand-launching-of-compfest-18-bridging-the-digital-skill-gap-and-building-adaptive-talents-in-the-8a1fca80fde3">Read English Version Here</a></p><p>Inovasi kecerdasan buatan (AI) dan evolusi teknologi yang berkembang dengan cepat menimbulkan <em>digital skill gap</em>, suatu krisis yang dihadapi dunia saat ini. Krisis ini berpotensi menghambat inovasi dan produktivitas di berbagai sektor. Akibatnya, timbul ketimpangan antara tingginya kebutuhan keahlian teknologi modern dengan minimnya talenta atau sumber daya yang ada.</p><p>Menjawab tantangan global yang urgensinya semakin meningkat, mahasiswa Fakultas Ilmu Komputer Universitas Indonesia menggelar <em>event</em> IT tahunan terbesar yang diadakan oleh mahasiswa, COMPFEST 18. Resmi dibuka dengan tema “Driving Infinite Growth to Enclose the Trench of Digital Skill Gap”, COMPFEST hadir sebagai katalis yang memastikan generasi muda Indonesia tidak hanya melek digital, tetapi juga adaptif dalam memimpin transformasi teknologi dan siap bersaing di era otomatisasi.</p><p>Menanggapi COMPFEST 18, Dekan Fasilkom UI, Prof. Dr. Eng. Wisnu Jatmiko, S.T., M.Kom, memberi dukungan penuh. Dalam sambutannya, beliau menekankan esensi COMPFEST yang tidak semata-mata program kerja rutin, namun merupakan sebuah warisan intelektual. “Melalui kompetisi, akademi, dan ruang eksplorasi yang telah disiapkan di COMPFEST 18, kita sedang berupaya melahirkan talenta-talenta digital yang tidak hanya mahir secara teknis, tetapi juga adaptif dan solutif dalam menjawab kebutuhan zaman,” tegasnya. Visi tersebut diselaraskan oleh Ketua BEM Fasilkom UI 2026, Makarim Zufar Prambudyo, yang menekankan pentingnya membekali generasi muda agar bertransformasi dari sekadar penonton menjadi penggerak solusi.</p><p>Melengkapi kedua dukungan tersebut, <em>Project Officer</em> COMPFEST 18, Khansa Dinda Arya Putri, ikut memaparkan pentingnya penyelenggaraan COMPFEST yang diharapkan mampu menutup jurang kesenjangan kemampuan digital. Menurutnya, acara tahunan yang sudah dimulai semenjak tahun 2009 menjadi wujud nyata kontribusi mahasiswa Fasilkom UI dalam memajukan ekosistem teknologi di Indonesia bagi masyarakat luas.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ln181HzZd4grBLZBd_Tojg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9Fw-RiwKn_4GowyZFFdrUw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d7abZgPlfMyMiZKWRLhXaA.png" /><figcaption>Sesi sambutan Grand Launching COMPFEST 18 oleh Prof. Dr. Eng. Wisnu Jatmiko, S.T., M.Kom (Dekan Fasilkom UI), Makarim Zufar Prambudyo (Ketua BEM Fasilkom UI 2026), dan Khansa Dinda Arya Putri (<em>Project Officer</em> COMPFEST 18). (Sumber: Dokumentasi)</figcaption></figure><p>Menariknya, bekal utama yang harus dimiliki untuk menghadapi gempuran teknologi saat ini ternyata bukan sekadar kemampuan menulis kode program. Topik ini dikupas tuntas dalam sesi Keynote Speech yang justru mematahkan stereotip lama tentang keahlian anak IT. Bapak Kristiyanto dari Google for Education Indonesia membuka sesi ini dengan menyoroti pergeseran nilai jual unik (<em>unique selling point</em>) seorang talenta. Di saat kecerdasan buatan sudah mampu mengambil alih tugas analitis dan pemecahan masalah, senjata utama yang dimiliki manusia mulai bergeser ke kemampuan komunikasi, presentasi, dan persuasi.</p><p>Melanjutkan pembahasan mengenai disrupsi oleh AI tersebut, pakar teknologi dan AI, Bapak Risman Adnan, Ph.D., memperkenalkan konsep AI <em>Native Engineering</em>. Alih-alih memandang kecerdasan buatan sebagai ancaman dalam merebut lapangan pekerjaan, beliau menegaskan kegunaan AI sebagai instrumen mutlak untuk mempercepat terciptanya inovasi. “AI bukanlah ancaman, tapi orang-orang yang tidak menggunakan AI akan sangat terancam dengan teknologi ini,” pesannya.</p><p>Sesi Keynote Speech ditutup oleh praktisi <em>cyber security</em>, Bapak Alfons Tanujaya, yang memberikan “sentilan tajam” mengenai pentingnya kelincahan dalam beradaptasi. Didasari atas kesuksesannya berkecimpung di dunia IT meskipun berlatar belakang pendidikan Ekonomi, beliau membuat analogi bahwa industri masa kini mirip seperti seleksi alam, “Bukan T-Rex yang bertahan sampai hari ini, yang <em>survive</em> adalah kecoa karena ia beradaptasi. Bukan yang paling kuat yang menang, tapi yang mampu menyesuaikan diri.” Lebih lanjutnya, beliau memberikan pesan bahwa di tengah ancaman kejahatan siber yang semakin canggih, talenta IT dituntut untuk menjadi garda terdepan sebagai pelindung ekosistem digital masyarakat.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pgBwktII1_0pZ564JacYBw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MCkIr2u9QfHUZ1frw0bGaQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Hmkw0U5vJWQFJHW075dgVA.png" /><figcaption>Jajaran pengisi sesi Keynote Speech: Bapak Kristiyanto (<em>Adaption Lead</em> di Google for Education Indonesia), Bapak Risman Adnan, Ph.D. (Pakar Teknologi &amp; AI), dan Bapak Alfons Tanujaya (IT &amp; <em>Financial Security Specialist</em> di vaksin.com). (Sumber: Dokumentasi)</figcaption></figure><p>Jika sesi Keynote Speech terkesan sangat konseptual dan sulit untuk dibayangkan, sesi selanjutnya, yaitu Talkshow bertopik “Reality Check: Bridging the Gap Between What You Know and What the Industry Demands”, akan membawa poin-poin tersebut ke realita lapangan. Dipandu oleh Ibu Kamila Alifia Imanuddin, S.Kom, M.T.I. sebagai moderator, sesi ini akan membedah kesenjangan antara teori di kampus dengan tuntutan di lingkungan kerja bersama Bapak Muhammad Mudrik, seorang <em>software engineer </em>di Flip, dan Bapak Abdul Hasan Zawawi, seorang AI &amp; <em>Machine Learning Engineer</em> di Prudential.</p><p>Dalam sesi <em>talkshow</em>, Bapak Hasan menyoroti perbedaan yang signifikan antara ekspektasi akademis dan industri. Di dunia perkuliahan, mahasiswa dinilai dari kemampuan mereka dalam menjawab soal-soal dengan benar. Akan tetapi, di dunia kerja, solusi yang dihasilkan harus cepat dan mampu membawa dampak yang nyata bagi bisnis perusahaan. Beliau juga mengingatkan mahasiswa untuk tidak hanya berfokus pada IPK, tetapi juga membangun portofolio lewat proyek-proyek dan mengasah <em>soft skills</em> yang tidak bisa ditiru mesin.</p><p>Di sisi lain, Bapak Mudrik menyoroti dilema yang dirasakan masyarakat akibat kemudahan teknologi saat ini. Walaupun AI memang dapat membantu dalam menuliskan kode atau menyelesaikan tugas, pemahaman konsep fundamental seperti logika dan struktur datalah yang justru membedakan seorang <em>engineer</em> yang andal dengan pengguna “jalan pintas” instan. “Jadikan AI sebagai asisten untuk belajar memahami konsep, bukan jalan pintas penyelesai tugas,” tekannya. Beliau menambahkan bahwa pada akhirnya, hal yang esensial dalam eksekusi proyek di bidang industri adalah kemampuan berkomunikasi dan berkolaborasi dalam tim, terlepas dari seberapa tinggi <em>hard skill</em> yang dimiliki oleh orang tersebut.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3nEX_ZLtxius4n04TdzJLw.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UdOUAHBOd6mqyLKSi3zXGg.jpeg" /><figcaption>Sesi Talkshow “Reality Check” bersama Ibu Kamila Alifia Imanuddin, S.Kom, M.T.I., Bapak Muhammad Mudrik (<em>Software Engineer </em>Flip), dan Bapak Abdul Hasan Zawawi (AI &amp; <em>Machine Learning Engineer </em>Prudential. (Sumber: Dokumentasi)</figcaption></figure><p>Seluruh ilmu dan <em>insight</em> yang berharga dari sesi Keynote Speech dan Talkshow pada dasarnya menjadi pembuka dari perjalanan COMPFEST 18. Untuk menjawab tuntutan nyata di dunia industri dan membekali generasi muda dengan amunisi yang diperlukan dalam menghadapi lingkungan kerja, COMPFEST 18 menyiapkan berbagai program unggulan.</p><p>Para peserta dapat mengasah kemampuan teknis dan manajerial mereka secara mendalam melalui empat pilar akademi utama, yaitu <em>Data Science Academy</em>, <em>Software Engineering Academy</em>, <em>User Experience Academy</em>, dan <em>Product Management Academy</em>. Selain itu, ruang untuk berinovasi juga disiapkan lewat kompetisi-kompetisi, mulai dari <em>Competitive Programming</em>, <em>Capture the Flag</em>, sampai <em>Business IT Case</em>.</p><p>Keseluruhan akademi dan kompetisi akan mencapai puncaknya pada <em>Main Event </em>yang menghadirkan pengalaman interaktif dan <em>Job Fair</em> untuk menghubungkan talenta dengan perusaahan impian. Oleh karena itu, pantau terus informasi lengkap mengenai pendaftaran dan jadwal acara melalui situs web resmi maupun media sosial COMPFEST agar tidak tertinggal dalam mengambil peran dalam transformasi digital saat ini! Mari bersama-sama persiapkan diri, asah potensi, dan jangan lupa untuk <em>enclose the gap!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8afe55bd1eae" width="1" height="1" alt=""><hr><p><a href="https://medium.com/compfest/grand-launching-compfest-18-menjembatani-digital-skill-gap-mencetak-talenta-adaptif-di-era-ai-8afe55bd1eae">Grand Launching COMPFEST 18: Menjembatani Digital Skill Gap, Mencetak Talenta Adaptif di Era AI</a> was originally published in <a href="https://medium.com/compfest">COMPFEST</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>