Building AI systems that think in feedback loops.
Data & AI product builder focused on LLM pipelines, autonomous agents, and product strategy. Ex-Cisco engineer, currently pursuing MSIM at UIUC.
Data & AI product builder focused on LLM pipelines, autonomous agents, and product strategy. Ex-Cisco engineer, currently pursuing MSIM at UIUC.
I translate technical complexity into products people actually use. Previously at Cisco Systems, I built ML pipelines and AI agents for enterprise security clients, shipping features to global production.
Currently an Associate Consultant at UIUC BIG, I design LLM-powered knowledge bases and own products end-to-end. I care about building reliable systems that solve real user problems.
LangChain, Claude API, OpenAI, RAG, Prompt Eng
dbt Core, Airflow, DuckDB, MinIO, Star Schema
Scikit-Learn, TensorFlow, Random Forest, K-Means
Tableau, Power BI, PRDs, Agile, Roadmapping
Led product development for an LLM-powered knowledge base for a B2B client. Authored PRDs, conducted user interviews, and prototyped RAG pipelines.
Shipped an AI-powered NLP packet analysis agent to global production. Managed 500+ Fortune 500 client engagements and processed 100K+ daily network events.
Focus on Data Engineering, AI Systems Evaluation, and Applied Analytics. Prior B.Tech from Manipal University (Minor in Data Science).
End-to-end agentic pipeline that classifies fault types using LangChain + NLP. Reached 80% accuracy on 10K+ captures.
LLM-powered retrieval system designed for a B2B healthcare client. Owned full lifecycle from PRD scoping to MVP prototype.
Cloud-portable ELT platform using PostgreSQL, MinIO, dbt Core, and DuckDB with automated data quality checks.
LangChain and FastAPI system that chunks/embeds PDFs and autonomously routes queries with a full observability layer.
End-to-end CSV analytics product. React + FastAPI architecture that auto-detects templates and generates AI insights.
Analyzed 1.3K records using Random Forest (R²=0.82) and K-Means to identify distinct health-risk demographic clusters.
ETL pipeline processing 1M+ records into a star schema data warehouse with 99.5% accuracy, visualized in Tableau.
ML classifier to detect malware via static analysis of PE headers. Random Forest achieved 99.6% accuracy on 51K samples.
Real-time CV system for driver fatigue. Achieved 90% accuracy processing 30 FPS video using TensorFlow and OpenCV.
Independent research on novel metaheuristic scheduling algorithms. Achieved 85% efficiency and 80% resource utilization.
Hybrid system using collaborative filtering on sparse user-item matrices built from implicit play count feedback.
Always open to discussing AI products, agentic workflows, or analytics engineering!