<?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 Nitin Mali on Medium]]></title>
        <description><![CDATA[Stories by Nitin Mali on Medium]]></description>
        <link>https://medium.com/@nitinmali_68448?source=rss-7a14164e17be------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*JcjZf-AZbWXOa2S4.</url>
            <title>Stories by Nitin Mali on Medium</title>
            <link>https://medium.com/@nitinmali_68448?source=rss-7a14164e17be------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 14:58:00 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@nitinmali_68448/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[Unlocking the Power of Skill Visualization in Recruitment: A Sankey Diagram Approach]]></title>
            <link>https://medium.com/@nitinmali_68448/unlocking-the-power-of-skill-visualization-in-recruitment-a-sankey-diagram-approach-5be4e182e602?source=rss-7a14164e17be------2</link>
            <guid isPermaLink="false">https://medium.com/p/5be4e182e602</guid>
            <category><![CDATA[plotly]]></category>
            <category><![CDATA[skills]]></category>
            <category><![CDATA[recruiting]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data-visualization]]></category>
            <dc:creator><![CDATA[Nitin Mali]]></dc:creator>
            <pubDate>Wed, 10 Jan 2024 08:24:52 GMT</pubDate>
            <atom:updated>2024-01-10T08:24:52.395Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*x7tjEmI7lMK8K9bbFX085Q.png" /></figure><p>Recruitment is an art finely tuned to the science of skill assessment. In the ever-evolving tech landscape, recruiters face the challenge of matching the multifaceted expertise of candidates to the dynamic needs of industry roles. This is where advanced visualization tools like Sankey diagrams come into play, transforming the recruitment process into a more data-driven, insightful endeavor.</p><p>The Sankey diagram above, intricately woven into the fabric of my professional portfolio at <a href="https://mroneai.com/">mroneai.com</a>, offers a panoramic view of my career’s skill trajectory. Here’s a breakdown of what this visualization encapsulates and how it can be a game-changer for recruiters:</p><h3>Visual Storytelling of Professional Growth</h3><p>The diagram charts a course through the professional landscape, illustrating the progressive development of skills acquired from one role to the next. It provides a storytelling canvas that narrates a candidate’s journey through various domains, skills, and competencies.</p><h3>High-Level Skill Categorization</h3><p>The horizontal bands represent different companies, colored uniquely to distinguish each tenure. Flowing from these are streams that converge into skill categories, showcasing areas of expertise such as Analytics, Machine Learning, and Business Strategy. This categorization allows recruiters to quickly assess a candidate’s areas of proficiency and how they’ve been applied across different job roles.</p><h3>Detailed Skill Insights</h3><p>On the far right, the streams bifurcate into specific skills, offering a granular view of the candidate’s abilities. From ‘Data Storytelling’ to ‘Gen AI and NLP’, each skill is a testament to tasks performed, challenges overcome, and knowledge applied. Recruiters can delve into the specifics, understanding not just the what, but the how and why of a candidate’s skill set.</p><h3>Benefits for Recruiters</h3><ul><li>Efficiency: At a glance, recruiters can map a candidate’s skill set against job requirements, saving time and resources in the selection process.</li><li>Precision: The detailed breakdown helps pinpoint the exact fit for specialized roles, ensuring that the right talent is considered for the right position.</li><li>Dynamism: It reflects a candidate’s adaptability and learning curve, essential qualities in the fast-paced tech industry.</li></ul><h3>Behind the Diagram</h3><p>For enthusiasts and fellow data scientists interested in the mechanics of this visualization, the detailed code is hosted on my <a href="https://github.com/nitzmali/portfolio">GitHub profile</a>. The diagram was crafted using Plotly’s robust visualization capabilities, integrated within a Streamlit application to create an interactive, user-friendly experience.</p><h3>Final Thoughts</h3><p>This Sankey diagram is more than just a visual treat; it’s a bridge connecting the dots of professional experience to potential opportunities. For recruiters, it’s a lens focusing on the essence of a candidate’s professional narrative, ensuring a symbiotic match between talent and enterprise.</p><p>To explore this visualization and for a detailed understanding of the skills mentioned, please visit the ‘WORK EXPERIENCE’ tab at <a href="https://mroneai.com/">mroneai.com</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5be4e182e602" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Predicting Job Titles, Skills, and Recommending Online Courses with Machine Learning]]></title>
            <link>https://medium.com/@nitinmali_68448/predicting-job-titles-skills-and-recommending-online-courses-with-machine-learning-0aad5d28a4f5?source=rss-7a14164e17be------2</link>
            <guid isPermaLink="false">https://medium.com/p/0aad5d28a4f5</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[generative-ai-tools]]></category>
            <category><![CDATA[career-transitions]]></category>
            <category><![CDATA[career-advice]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Nitin Mali]]></dc:creator>
            <pubDate>Thu, 04 Jan 2024 07:09:02 GMT</pubDate>
            <atom:updated>2024-01-04T07:09:02.370Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1021/1*pZ0kQ4rMiqpDeUIfZYJFgw.png" /><figcaption>Schematic Representation of the Process</figcaption></figure><p>Introduction:</p><p>In today’s fast-paced world, having the right skills and finding the right job can be a challenging task. With the advancements in machine learning, we can now leverage the power of data to predict job titles, identify required skills, and even recommend relevant online courses to bridge the skill gap. In this article, we will explore a Python script that utilizes machine learning models to achieve these tasks. Let’s dive in!</p><p>Section 1: Preparing the Environment</p><p>Before we jump into the code, let’s set up our environment by installing the necessary libraries and modules. We will need pandas, TensorFlow, and transformers libraries for data manipulation, deep learning, and natural language processing tasks.</p><p>Section 2: Job Prediction Model</p><p>To predict job titles, we will use a deep learning model that consists of an encoder and a decoder. The encoder model encodes the input sequence, and the decoder model generates the output job title. We will also need a tokenizer to process the text data.</p><p>The code snippet provided defines a class called loadjobModels that handles loading the job prediction models and tokenizer. The load_m method loads the models and tokenizer from their respective files. The decode_sequence_job method decodes the input sequence and generates the output job title using the loaded models and tokenizer.</p><p>Section 3: Skills Prediction Model</p><p>To predict skills, we will again use a deep learning model with an encoder and a decoder. Similar to the job prediction model, we need to load the models and tokenizer for skills prediction. The loadskillsModels class handles loading the models and tokenizer, while the decode_sequence_skills method decodes the input sequence and generates the output skills.</p><p>Section 4: Text Preprocessing and Embeddings</p><p>Before applying the models, we need to preprocess the text data and convert it into numerical embeddings. In our code, we use the BERT model for text encoding. The get_embedding function preprocesses the text and generates embeddings using the BERT model. The batch_encode function batches the text data and generates embeddings in batches for efficiency.</p><p>Section 5: Job Prediction</p><p>Now that we have our job prediction model and text embeddings, we can predict the next job title based on the user’s current job and skills. The predict_job function takes a user dataframe as input, generates the embeddings for the text columns, and predicts the next job title using the job prediction model.</p><p>Section 6: Skills Prediction</p><p>Similarly, we can use the skills prediction model to predict the next set of skills based on the user’s current job and skills. The predict_skills function takes a user dataframe as input, generates the embeddings for the text columns, and predicts the next set of skills using the skills prediction model.</p><p>Section 7: Course Recommendation</p><p>To recommend relevant online courses, we calculate the similarity between the user’s predicted skills and the courses’ descriptions. The course_metrics class provides methods for calculating course scores and selecting relevant courses. The calculate_course_score method calculates the score of a course based on skill match, popularity, and rating. The calculate_similarity_score method calculates the cosine similarity between the skill gap and course descriptions. Finally, the select_relevant_course method selects the top course based on the calculated scores and similarity.</p><p>Section 8: Putting It All Together</p><p>In the main function, we bring all the pieces together. We load the necessary datasets, iterate the job and skills prediction until we reach the final job or a maximum number of iterations. We then call the course recommendation function to predict relevant courses based on the predicted skills. The results are stored in a dataframe and returned as the output of the script.</p><p>Conclusion:</p><p>In this article, we explored a Python script that utilizes machine learning models to predict job titles, identify required skills, and recommend relevant online courses. By leveraging the power of data and deep learning, we can empower individuals in their career journey. Whether you’re an aspiring professional or looking to make a career transition, these predictive models can provide valuable insights and guidance. So, go ahead and give it a try! Happy coding and learning!</p><p>Acknowledgements:</p><p>This project wouldn’t have been possible without the contributions of various open-source libraries and datasets. We would like to thank the developers and contributors of pandas, TensorFlow, transformers, and other related libraries for making their work available to the community.</p><p>References:</p><ol><li>Full Code: <a href="https://github.com/nitzmali/job_transition_pathway">https://github.com/nitzmali/job_transition_pathway</a></li><li>Pandas Documentation.</li><li>TensorFlow Documentation.</li><li>Transformers Documentation.</li></ol><p>#CareerAdvice#JobSearch#MachineLearning#DataScience#ArtificialIntelligence#DeepLearning#SkillsDevelopment#OnlineLearning#ProfessionalDevelopment#Technology#CareerTransition#JobMarket#LinkedIn#CareerGuidance#SkillGap#Education#CareerTips#JobOpportunities#TechIndustry#PersonalGrowth</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0aad5d28a4f5" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>