| β‘ 40% Latency Cut | π― 84% RAG Precision | π€ 65% Manual Work Eliminated | π₯ 500+ Users |
|---|---|---|---|
| NL-to-SQL on AWS Lambda | Redis + Cohere Rerank | LangGraph Vertical Agent | Taylor Chrome Extension |
π₯ EnterpriseClaw β Production Agentic Framework
Deterministic agentic framework with LangGraph state machines, human-in-the-loop guardrails, and MCP tool routing. Built for enterprise environments where agent failures have real consequences.
LangGraph PydanticAI MCP HITL Python
π§ Vertical AI Agent β Live Demo β
Manager-Specialist multi-agent system built with PydanticAI + MCP. Orchestrates Gmail, Drive, Jira, and Calendar in stateful multi-step workflows β cutting manual task time by 65%.
PydanticAI MCP Multi-Agent FastAPI Python
Vertical AI agent automating financial data extraction from unstructured documents, cross-source research synthesis, and investment memo generation β 10Γ faster than manual analysis.
LangGraph Advanced RAG IDP Python LLMs
ποΈ JerseySTEM SQL Agent (Private)
Production NL-to-SQL agent deployed on AWS Lambda. Reduced query latency by 40% with intelligent caching and schema-aware retrieval, serving the full JerseySTEM org.
AWS Lambda NL-to-SQL LangChain PostgreSQL Redis
Document intelligence service with structure-preserving chunking and section-aware retrieval. Achieved 84% precision using semantic search + Cohere reranking over dense enterprise docs.
RAG Cohere Rerank FastAPI Vector Search Python
ποΈ GemmA.I β On-Device Vision Assistant (iOS)
Accessibility-first iOS app delivering real-time scene understanding for visually impaired users β fully on-device, zero cloud calls, complete privacy.
Swift iOS On-Device ML Computer Vision Edge AI
βοΈ Taylor β AI Career Stylist
AI agents serving 500+ users. Tailors resumes and cover letters per job description with one click β ATS-optimized, with application streak tracking and a live dashboard.
LLM Agents Chrome Extension NLP HTML/JS
π NJ Transit Assistant β π₯ HackRU 2024
Scalable RESTful API for real-time transit predictions. Redis caching + PostgreSQL + load balancing for a fault-tolerant, low-latency distributed system under event load.
FastAPI PostgreSQL Redis Python Distributed Systems
Agentic AI
Infrastructure
AI/ML
π₯ HackRU 2024 β 1st Place Β |Β π IEEE Published Research Β |Β π MS Data Science @ Rutgers Β |Β π 500+ Production Users