Insurance sector - Generative AI developer• Client - Open Enterprise AI - needed a platform where users can upload their medical documents and the AI will judge if the user is a good candidate.• Worked on developing Generative AI systems using RAG (Retrieval-Augmented Generation) to overcome context limitations of LLM. Developed multimodal generative AI system to understand texts from PDFs of medical files.Medical device compliance AI – Prompt engineer • Client wanted to automate their device compliance verification.• Worked on prompt engineering to fetch the right data (documents were 10 – 300 pages), according to the user’s natural language query from the SQL database. Used Python to build the backed framework and used GPT 4 LLM.Retail/ Grocery sector – Generative AI engineer• Client – Bhargavi Hypermarket - client wanted to migrate their huge database into cloud with AI based analytics with Natural Language interaction.• Worked on cloud database creation, analytics engine was built with Python, NumPy and Pandas framework. Finally used vector database and RAG system for GPT-4 LLM to fetch the data query from.

Google and ONDC hackathon – Gen AI Legal compliance advisor• Tackled the compliance problem by building a Gen AI chatbot that is fed all the laws and regulations of India and can hence advice on complex legal framework for any product in seconds accurately like a lawyer.Customer care AI chatbot – Generative AI engineer• Client Sansav group – Built a Generative AI chatbot on GPT-3 API which fetched the information asked by the customers from their live database eliminating the need for dedicated customer care team.Vertiko.io – Backend System developer • Built the entire backend system of Vertiko.io and implemented Generative AI video chatbot to increase their website sales conversion rate.

• Kandity, AI based sports social media app promoting engagement for fans.• Developed AI algorithms to analyse and track user behavior, identifying preferred content and providing personalized recommendations. Built light weight backend to keep the costs low as the startup was bootstrapped