<?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[Stories by Aysell Unlu on Medium]]></title>
        <description><![CDATA[Stories by Aysell Unlu on Medium]]></description>
        <link>https://medium.com/@aysell.unlu?source=rss-4fae10fb7400------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*pehcLeyOucevrYQZNulYJw.png</url>
            <title>Stories by Aysell Unlu on Medium</title>
            <link>https://medium.com/@aysell.unlu?source=rss-4fae10fb7400------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 07:41:56 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@aysell.unlu/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Building a Lean Lead Generation Engine with Mastra and Groq]]></title>
            <link>https://medium.com/@aysell.unlu/building-a-lean-lead-generation-engine-with-mastra-and-groq-58d7094f3004?source=rss-4fae10fb7400------2</link>
            <guid isPermaLink="false">https://medium.com/p/58d7094f3004</guid>
            <category><![CDATA[mastra-ai]]></category>
            <category><![CDATA[groq]]></category>
            <category><![CDATA[lead-generation]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[ai-agent]]></category>
            <dc:creator><![CDATA[Aysell Unlu]]></dc:creator>
            <pubDate>Fri, 20 Feb 2026 20:42:30 GMT</pubDate>
            <atom:updated>2026-02-20T20:42:30.458Z</atom:updated>
            <content:encoded><![CDATA[<p>Automating lead generation often involves complex scraping infrastructures. However, the <strong>mastra-101-lead-gen</strong> project demonstrates a more streamlined, “native” approach. By combining the <strong>Mastra</strong> framework’s built-in tool capabilities with <strong>Groq’s</strong> high-speed inference, we can create a powerful lead discovery engine that is both lightweight and highly effective.</p><h3>The Architectural Approach: Keep It Lean</h3><p>Unlike systems that rely on heavy external crawlers, this project focuses on <strong>Search Integration</strong> and <strong>Intelligent Filtering</strong>. The goal is to identify potential business leads directly through search intent and model-driven qualification.</p><h3>1. Mastra Framework as the Native Orchestrator</h3><p>In this project, Mastra isn’t just a wrapper; it’s used to define custom <strong>Tools</strong> that interact with Search APIs. Mastra’s ability to handle tool-calling with strict TypeScript schemas ensures that the lead data (company names, descriptions, links) remains structured and reliable throughout the process.</p><h3>2. High-Speed Qualification with Groq</h3><p>The bottleneck in lead generation is usually the “evaluation” phase — deciding if a search result is actually a relevant lead. By using <strong>Groq</strong>, the agent can process and analyze multiple search results in parallel with near-zero latency. This allows for real-time lead scoring without the traditional “loading” delays seen in GPT-4 based agents.</p><h3>3. Integrated Search Strategy</h3><p>The repository leverages direct Search API integrations to find companies within a specific niche. The agent acts as a filter:</p><ul><li><strong>Search:</strong> It queries the web for targeted industries.</li><li><strong>Filter:</strong> It uses the LLM to scan snippets and metadata to discard irrelevant results.</li><li><strong>Structure:</strong> It outputs a clean, actionable list of leads.</li></ul><h3>How the Pipeline Works</h3><ol><li><strong>Targeting:</strong> The user defines a niche (e.g., “Software Agencies in Istanbul”).</li><li><strong>Autonomous Search:</strong> The Mastra agent uses its search tool to gather the top organic results.</li><li><strong>Lead Scoring:</strong> Groq analyzes the context of each result — checking for service relevance, company size, or any specific criteria defined in the system prompt.</li><li><strong>Data Export:</strong> The final output is a curated list of high-intent prospects.</li></ol><h3>Why This Approach Matters</h3><ul><li><strong>Zero-Bloat:</strong> By avoiding external web-crawling services where they aren’t strictly necessary, you reduce both cost and architectural complexity.</li><li><strong>Speed-First:</strong> The combination of native Mastra tools and Groq makes this one of the fastest lead-gen prototypes available.</li><li><strong>Precision:</strong> The agent focuses on search intent, which often yields higher-quality business results than “blind” web scraping.</li></ul><p><a href="https://github.com/aysellunlu/mastra-101-lead-gen">https://github.com/aysellunlu/mastra-101-lead-gen</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=58d7094f3004" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Developing an Agentic Research Hub with Mastra, Groq, and Firecrawl]]></title>
            <link>https://medium.com/@aysell.unlu/developing-an-agentic-research-hub-with-mastra-groq-and-firecrawl-f2903dabe9a1?source=rss-4fae10fb7400------2</link>
            <guid isPermaLink="false">https://medium.com/p/f2903dabe9a1</guid>
            <category><![CDATA[ai-agent]]></category>
            <category><![CDATA[typescript]]></category>
            <category><![CDATA[ai-solution]]></category>
            <category><![CDATA[mastra-ai]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Aysell Unlu]]></dc:creator>
            <pubDate>Fri, 20 Feb 2026 20:34:06 GMT</pubDate>
            <atom:updated>2026-02-20T20:34:59.344Z</atom:updated>
            <content:encoded><![CDATA[<p>We are moving past simple chat interfaces into the era of autonomous agents — systems that can research, filter internet data, and provide structured reports independently. This article breaks down the architecture and logic of the <strong>mastra-agentic-research-hub</strong> project.</p><h3>Core Components of the Stack</h3><p>The project is built on three pillars, each solving a specific problem in the agentic workflow:</p><ol><li><strong>Mastra Framework (The Orchestrator):</strong> Mastra acts as the central nervous system. It provides the structure for defining Agents, Tools, and Workflows. Being TypeScript-based and modular, it ensures that the development process is standardized and scalable.</li><li><strong>Groq (The Inference Engine):</strong> Speed is critical for agents. Unlike traditional providers, Groq offers ultra-low latency. Since an agent often goes through multiple “reasoning steps” before finishing a task, Groq reduces the total processing time from minutes to seconds.</li><li><strong>Firecrawl (Data Acquisition):</strong> This is how the agent “reads” the web. Firecrawl bypasses the noise of modern websites (scripts, ads, complex HTML) and converts content directly into clean <strong>Markdown</strong>. This ensures the agent receives high-quality data, which leads to high-quality output.</li></ol><h3>How the System Works</h3><p>The research process follows a logical sequence managed by a Mastra workflow:</p><h3>1. Planning and Query Expansion</h3><p>When a user submits a topic, the agent (powered by Groq) analyzes the intent. It breaks down the high-level request into specific search queries and identifies the information gaps it needs to fill.</p><h3>2. Intelligent Data Retrieval</h3><p>The agent triggers Firecrawl to fetch data from the web. Unlike basic scrapers, Firecrawl handles JavaScript-heavy sites effectively, ensuring the agent doesn’t miss information locked behind modern web architectures.</p><h3>3. Synthesis and Reporting</h3><p>Mastra orchestrates the final step where the gathered data is processed. Groq cross-references the findings, removes inconsistencies, and synthesizes the information into a structured, professional report.</p><h3>Why This Architecture Works</h3><ul><li><strong>Efficiency:</strong> Groq eliminates the bottleneck of slow LLM responses during multi-step reasoning.</li><li><strong>Data Integrity:</strong> Firecrawl ensures the model isn’t confused by “dirty” HTML data.</li><li><strong>Modularity:</strong> Using Mastra makes it easy to swap models or add new tools (like a PostgreSQL database for memory) without rewriting the core logic.</li></ul><h3>Conclusion</h3><p>The <strong>mastra-agentic-research-hub</strong> is a clean blueprint for modern AI applications. It focuses on performance and reliability rather than hype. By combining a solid orchestration framework with high-speed inference and precise data scraping, you can build production-ready agents for market research, technical analysis, or lead generation.</p><p><a href="https://github.com/aysellunlu/mastra-agentic-research-hub">https://github.com/aysellunlu/mastra-agentic-research-hub</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f2903dabe9a1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[React Native Nedir]]></title>
            <link>https://medium.com/@aysell.unlu/react-native-nedir-01daa10b076d?source=rss-4fae10fb7400------2</link>
            <guid isPermaLink="false">https://medium.com/p/01daa10b076d</guid>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[mobile-app-development]]></category>
            <category><![CDATA[reactjs]]></category>
            <category><![CDATA[react]]></category>
            <category><![CDATA[react-native]]></category>
            <dc:creator><![CDATA[Aysell Unlu]]></dc:creator>
            <pubDate>Fri, 03 Nov 2023 12:37:15 GMT</pubDate>
            <atom:updated>2023-11-03T12:37:15.847Z</atom:updated>
            <content:encoded><![CDATA[<h3>React Native Nedir ?</h3><p><strong>1.1 React Native Nedir ?</strong></p><p>React Native, cross platform mobil uygulama geliştirmede kolay kullanıma sahip, Javascript dili ile geliştirilmeye açık bir framework’tür. Web uygulamalarında kullanılan React.js kütüphanesi temellidir.</p><p><strong>1.2 React Native Kullanan Firmalar</strong></p><p>Mobil uygulama geliştirmeye yönelik olan React Native çerçevesi, 2013 yaz aylarında Facebook tarafından geliştirilmeye başlanmıştır. İlk tanıtımı Ocak 2015’te yapılan React Native, Mart 2015’te GitHub’da açık olarak paylaşılmış ve kullanılabilir hale getirilmiştir.</p><ul><li>Facebook</li><li>Instagram</li><li>Whatsapp</li><li>Netflix</li><li>Discord</li><li>New York Times</li><li>Skype</li><li>Tesla Motors</li><li>Pinterest</li><li>Trendyol</li><li>Bloomberg</li></ul><p>gibi binlerce uygulama React Native ile geliştirilmiştir.</p><p><strong>1.3 Cross Platform Uygulama Geliştirme</strong></p><p>Cross Platform ( Çapraz Platform ), tek bir yerde yazılan kodların her platformda çalışabilir olmasıdır.</p><p>React Native de cross platform uygulama geliştirmeye olanak tanır. Yani, bu framework ile hem Android hem de iOS için çalışan uygulamalar oluşturulabilir. iOS için Objective-C / Swift , Android için Java / Kotlin yerine React Native framework’ü içerisinde Javascript ile mobil uygulama geliştirmek mümkündür.</p><p>Kendine özgü bir mobil uygulama uzantısı bulunmamaktadır. Android için <strong>.apk</strong> , iOS için <strong>.ipa</strong> uzantısı kolayca yaratılabilir.</p><p><strong>1.4 Native Code, JS Code ve React Native İlişkisi</strong></p><p>React Native, mobil uygulamalar geliştirmek için kullanılan bir framework’tür. React Native, hem “Native Code” (doğal kod) hem de “JavaScript Code” (JavaScript kodu) kullanarak çalışabilir.</p><p><strong>Native Code (Doğal Kod):</strong> Native Code, bir platformun (örneğin, Android veya iOS) doğal dilinde yazılan kodu ifade eder. Android uygulamaları Java veya Kotlin diliyle, iOS uygulamaları ise Swift veya Objective-C diliyle yazılır. Bu diller, ilgili platformlar tarafından tam olarak desteklenir ve doğrudan cihazın işletim sistemine uygun olarak çalışır.</p><p><strong>JavaScript Code (JavaScript Kodu): </strong>JavaScript Code, web geliştirme için kullanılan yaygın bir programlama dilidir. React Native, JavaScript Code ile çalışabilen bir mobil uygulama geliştirme çerçevesidir. React Native, JavaScript tarafından çalıştırılan bir JavaScript motoru içerir ve bu motor, uygulamanın temel mantığını ve işlevselliğini yönetir.</p><p><strong>React Native: </strong>React Native, bu iki dünyayı birleştiren bir yapıdır. React Native ile geliştirilen uygulamalar, arka planda doğal kod (Native Code) ile çalışan bileşenlerle etkileşimde bulunabilir. Örneğin, bir React Native uygulaması, kullanıcı arayüzü bileşenlerini (örneğin, düğmeler, metin kutuları) JavaScript Code ile oluşturabilir ve yönetebilirken, kamera veya sensörler gibi doğrudan donanım erişimine ihtiyaç duyulan yerlerde Native Code kullanılabilir. React Native, JavaScript Code ile yazılan uygulamanın, platformun doğal koduyla etkileşimde bulunmasına olanak tanır.</p><p>Bu durum, React Native’in popülerliğini artıran bir faktördür, çünkü geliştiriciler, tek bir kod tabanı üzerinde çalışarak hem Android hem de iOS platformlarına yönelik uygulamalar geliştirebilirler. Bu, geliştirme süresini kısaltabilir ve kaynakları daha verimli kullanmanıza yardımcı olabilir.</p><p>React Native, uygulama ve cihaz arasında bir köprü davranışı sergiler. Uygulama için kullanılan bileşenler cihazlara özgü karşılıklara dönüştürülür. Örneğin React Native içerisinde oluşturulan &lt;Text&gt; bileşeni Android için &lt;TextView&gt;, iOS için &lt;UITextView&gt; bileşenlerinde rollerine devam ederler. Burada React içerisinde Javascript ile yazılan kodlar Babel aracılığı ile Native dile ( Java, Kotlin, Objective-C, Swift…) çevrilmektedir.</p><p><strong>1.5 React Native ile Uyumlu Backend Servisleri</strong></p><p>Frontend geliştirme framework’ü olan React Native, destek ekibinin geliştirmeleri sayesinde çok fazla backend servisi tarafından desteklenmektedir.</p><ul><li>MongoDB,</li><li>SQLite,</li><li>Firebase,</li><li>GraphQL,</li><li>NodeJS ve daha fazlası…</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=01daa10b076d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Introduction]]></title>
            <link>https://medium.com/@aysell.unlu/introduction-48b49be367b4?source=rss-4fae10fb7400------2</link>
            <guid isPermaLink="false">https://medium.com/p/48b49be367b4</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[design]]></category>
            <category><![CDATA[front-end-development]]></category>
            <category><![CDATA[introduction]]></category>
            <category><![CDATA[engineering]]></category>
            <dc:creator><![CDATA[Aysell Unlu]]></dc:creator>
            <pubDate>Wed, 07 Dec 2022 20:14:55 GMT</pubDate>
            <atom:updated>2022-12-07T20:21:22.725Z</atom:updated>
            <content:encoded><![CDATA[<p>Hi there,</p><p>I have plan to write about my ideas, my perspectives, some development informations and other things in Turkish and English. I hope it will be an enjoyable read for you.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=48b49be367b4" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>