May 2025 – Jul 2025 · 2 mos · Remote · Remote

• At Pascal AI Labs, I am developing a comprehensive LLM evaluation platform that enables finance domain experts to conduct efficient prompt experimentation and systematic analysis of LLM responses against internal golden datasets, significantly reducing their manual effort and saving considerable time. I have designed and built both the user interface and backend infrastructure, creating intuitive comparison tools that allow domain experts to easily evaluate LLM outputs against golden answers and historical experiments, enhancing team efficiency and collaboration. Additionally, I have assisted Pascal AI in benchmarking their fintech software against the industry-standard FinBench dataset to establish baseline performance metrics, and through research, analysis, and implementation of system enhancements, I have improved normal RAG accuracy by 11% and advanced RAG accuracy by 10%, contributing to the company's performance in the fintech AI space.

Dec 2024 – Present · 1 yr 3 mos · Remote · Remote

• Writes articles on topics related to Machine Learning. More than 15k views on Medium Articles.
• Contributing author at Publications namely - Towards Data Science, Towards AI, AI Advances.

Dec 2024 – Mar 2025 · 3 mos · Washington DC · Remote

• As a Machine Learning Intern at JobTwine, I developed the Playbook feature, an intelligent interview creation tool that automates customized round and question generation using OpenAI APIs and LangChain. I built scalable APIs with a "min-to-max" approach, enabling users to create interviews from job titles to detailed specifications, improving customization by 60%. Leveraging RAG with PostgreSQL and Redis, I optimized skill retrieval and interview configuration management, reducing update times by 50%. Collaborating closely with the CTO, I refined product requirements and solved production issues, driving a 40% boost in user engagement and product quality (more details in resume above)

Jan 2024 – Jul 2024 · 6 mos · Brentwood, California, United States · Remote

• During my internship, I developed an AI-powered conversational training platform for 5-star hotel staff, designed to simulate realistic customer interactions and provide real-time feedback. The platform served as a training tool to enhance staff knowledge and professionalism across key hotel services like dining, rooms, and amenities etc. By leveraging openAI models, RAG-based search, and parallel processing, the system improved training accuracy by 50% and reduced latency by 30%, ensuring efficient and effective staff evaluation and development (more details in resume above)

Oct 2022 – Oct 2022 · 0 mo

• Event coordinator for Fashion Show Mr and Ms Thomso.
• A two day event including prelims and grand finale where around 100 participants participated.