• Developer 1

  July 2026 – Present

  • Analyzed and replaced autonomous LLM tool calls with direct API calls, reducing unnecessary model invocations and enabling 15% cost savings.
  • Implemented a cron job for the Mute Queue to remove expired Redis entries based on timestamps and update Langfuse traces with execution spans, contributing to an 8% reduction in LLM costs.

• Developer Intern

  Jan 2026 – June 2026

  • Developed a no-code platform for building, publishing, and managing AI agents.
  • Developed MCP tools to query a knowledge graph built on enterprise ticket data, enabling dynamic retrieval of historical data and structured insights.

Research Intern - Machine Learning

• Built reinforcement learning (RL) models for autonomous drone navigation using Proximal Policy Optimization (PPO), focusing on obstacle avoidance and goal-directed control in simulated environments.
• Developing predictive machine learning models to estimate battery State of Charge (SoC) with improved accuracy and robustness.

AI Engineer Intern

• Designed and deployed RESTful APIs for a compliance platform, enhancing performance and reducing response latency by over 40%.
• Developed a RAG-based chatbot leveraging hybrid retrieval techniques to deliver precise responses for ABPI case and clause-related queries, improving overall information accuracy.

Research Intern

• Built a RAG pipeline using LLMs to extract key symptoms from patient data and answer clinical queries with health reports as a knowledge base, leveraging a FAISS vector database to reduce hallucinations by over 50%.
• Developed a live streaming speech-to-text application using open-source ASR models, enhancing accuracy by 80%.