• Designed and implemented a scalable semantic search pipeline leveraging code parsing and vector embeddings with OpenAI LLMs to enhance the contextual understanding of extensive codebases across multiple projects.
• Developed an intuitive user interface using React that delivers traceable file-linked answers to developer queries in natural language, resulting in a 30% decrease in average onboarding ramp-up time for new engineers.
• Integrated GitHub OAuth with scoped permissions and multi-user workspace logic to facilitate secure collaboration among engineering teams while simplifying access management across diverse enterprise environments.
• Automated data ingestion processes from JavaScript and Python Git repositories which significantly reduced setup time while ensuring consistent indexing for AI-driven querying capabilities across applications.
• Proposed an innovative roadmap for fine-tuning LLMs and assessed advancements such as retrieval-augmented generation to potentially decrease hallucination rates by up to 40%, enhancing overall model reliability.
• Product developed under tech accelerator OS Labs (opensourcelabs.io)
Skills: Full Stack Development, Semantic Search, React.js, Node.js, OpenAI APIs