Sahil Kumar Nayak

I am a 2024 chemical engineering graduate from IIT Roorkee. During college time, I gained hands-on experience in designing equipment for the plant reactor, combustor, cyclone separator, riser, etc. Along with academics, I was also involved in technical fields like C++, Python, and SQL. I am also efficient in building AI chatbots and machine learning models. I did an internship at Futures First as a Market Analyst. I have worked freelancing at Outlier as a Prompt Engineer.

In my current role as a Research Associate AI at Keywords Studios, I have gained substantial experience in developing and deploying machine learning models, particularly in the context of large language models (LLMs). My primary focus has been on enhancing model performance through advanced prompt engineering techniques. For instance, I designed and implemented Chain of Thought (CoT) prompts that significantly improved reasoning capabilities, resulting in a 10% increase in performance accuracy. This involved not only crafting effective prompts but also understanding the underlying model architecture and how different inputs affect outputs. I also engineered prompt-based training workflows that optimized AI model retrieval performance across diverse web content, leading to a 25% boost in task efficiency. This experience has equipped me with a solid foundation in the end-to-end process of machine learning, from data preprocessing to model training and evaluation. Additionally, I have collaborated closely with software engineers to ensure seamless integration of these models into production environments. This collaboration has taught me the importance of documentation and communication in the deployment process, as it ensures that all stakeholders are aligned and that the models function as intended in real-world applications. My experience in using tools like TensorFlow and Scikit-Learn has further enhanced my ability to develop robust machine learning solutions.

Recently, I have built a RAG-powered AI chatbot that evolves langchain for building a pipeline, chroma database for storing vectorized data, and streamlit for streaming the result.

Overall, I am confident in my ability to contribute effectively to your team by leveraging my skills in machine learning model development and deployment.