Aug 2024 – Present · 1 yr 10 mos · Gurugram, Haryana, India · Hybrid

• 1. GenAI Chatbot for Defect Search & Analysis
• Technologies Used: Python, GenAI, Azure OpenAI, MongoDB, Flask, ChromaDB, Linux
• Project Overview:
• I worked as a Backend Developer on a GenAI-powered chatbot application designed to enhance defect searchability within the company's internal search engine. The existing system retrieved defect details from a database, but developers faced challenges in finding relevant defects and understanding the problem and its solution efficiently.
• To solve this, I built a GenAI chatbot using the Retrieval-Augmented Generation (RAG) approach. The system stores defect details in a vector database (ChromaDB) and processes user queries with contextual understanding. By leveraging cosine similarity-based search, the chatbot provides highly relevant results with clear explanations, making defect resolution faster and more intuitive.
• 2.Effort Estimation Chatbot for 3P Software Vulnerabilities
• Technologies Used: Python, GenAI, Azure OpenAI, MongoDB, Flask, ChromaDB, Linux
• Project Overview:
• As a POC-level project, I developed a GenAI-powered chatbot to automate effort estimation for fixing vulnerabilities in third-party (3P) software.
• In my company, different customers use various 3P software versions in their products. Keeping track of vulnerabilities and estimating the effort required to fix them was a manual and time-consuming process. To solve this, I built a chatbot application using the RAG approach.
• The chatbot:
• ✔ Analyzes user queries and extracts relevant context
• ✔ Searches for affected 3P software from a custom-built dataset containing 3P software details mapped to customers and product versions
• ✔ Identifies vulnerabilities
• ✔ Displays results in a tabular format, showing:
• Vulnerable 3P software
• Associated product/module
• CVE ID of vulnerabilities
• Estimated effort (man-days) required to fix the issues

Feb 2024 – May 2024 · 3 mos · Gurugram, Haryana, India · On-site