I build AI agents for enterprise teams, and the part I enjoy most is the messy middle, where a model meets real data, real users, and real constraints. I started out building data pipelines at Jio, which taught me to care about clean data, good tests, and systems that hold up. During my master's at NC State I got hooked on LLMs, and today I work as a Software Engineer on an AI platform at Infosys, building multi-agent workflows, RAG pipelines, and the evaluation and guardrails that make them reliable.
Infosys Limited, Raleigh, NC, USA
Architected an AI-powered analytics platform serving 1,200+ business users across 12 enterprise dashboards, with multi-agent workflows built on LangGraph and Agno, conversational analytics on Databricks Genie, and tools shared through an MCP hub. Cut hallucinations 44% and raised agent task success 35% with an LLM-as-a-judge evaluation pipeline in LangSmith, and reduced response time from 10–12s to 6–7s with parallel tool calls and query routing.
North Carolina State University, Raleigh, NC, USA
Mentored 148 graduate students on database course projects, guiding schema design, SQL optimization, and implementation, and evaluated and graded projects through live demonstrations.
U.S. Department of Agriculture, Albany, CA, USA
Built a GPT-4 RAG pipeline (LangChain, OpenAI embeddings, FAISS) to extract traits and QTLs from 36 genetics papers, achieving 97% paper categorization and 80% trait extraction. Built a genomic data search portal (Flask, JavaScript, MongoDB, Docker) that cut manual retrieval effort by 88%.
Reliance Jio Platforms Limited, Mumbai, India
Built forecasting pipelines with Azure ML, Databricks, and Azure Data Factory that reduced warehouse stockouts 25% across 20+ regions; led an address intelligence platform processing 8M+ delivery records that improved delivery accuracy 32%; and engineered 12+ production ETL pipelines (Kafka, Hive, MongoDB, SAP HANA) with CI/CD on Docker, Kubernetes, and Azure DevOps.
North Carolina State University, Raleigh, NC, USA
GPA 3.67/4.0. Algorithms and Data Mining, Neural Networks and Deep Learning, Software Engineering, Cloud Computing.
Land Registration System Using Block-chain
Proposed blockchain based land registration system which provides a transparent, secured, and decentralized method for execution of transactions between the participants by employing the concept of hyper ledger.
University Of Mumbai, India
GPA 9.17/10. Data Structures, Machine Learning, Natural Language Processing, Artificial Intelligence, Statistics
Issued By: Infosys Foundation USA - Best Use of AI
Issued On: March 2025
Issued By: North Carolina State University
Issued On: April 2024
Working on agents, RAG, or MCP, or hiring for an AI engineering role? Email is the fastest way to reach me.