Projects

Production-grade AI systems, multi-agent pipelines, RAG architectures, and ML projects.

Evidence-Grounded Civic Intelligence Platform
Featured
RAG/KG

Hybrid and graph RAG over ~1,600 civic meetings, with claim-level citation validation and enforced abstention when evidence is insufficient.

Python
FastAPI
MongoDB
Neo4j
Weaviate
+6

Claim-level citation validation + enforced abstention across ~1,600 meetings

AI Evaluation & Regression Harness
Featured
Evaluation

A versioned evaluation harness measuring retrieval, grounding, attribution, conversation and security behavior — used as a release gate rather than a one-off report.

Python
Evaluation Pipelines
Langfuse
OpenTelemetry
Retrieval Metrics
+1

Golden datasets & hard negatives as a versioned release gate

Multi-County Public Records Data Platform
Featured
Data Engineering

A canonical MongoDB platform unifying 1.7M+ public-record profiles and ~4M historical participation records from incompatible county schemas.

Python
MongoDB
PyMongo
ETL
Identity Resolution
+1

1.7M+ profiles · 20,060 new profiles reconciled, 1 ambiguous identity isolated

Geospatial Field Operations Application
Featured
Web Dev

A Next.js and MongoDB geospatial field app using GeoJSON, 2dsphere indexing and live device location to surface the nearest eligible records with contact and navigation actions.

Next.js
TypeScript
MongoDB
GeoJSON
Google Maps
+1

Live geospatial nearest-record search with real-time device location

MCP × A2A Hackathon — Multimodal Agent System
Featured
Agents & MCP

A multimodal autonomous agent system for complex task automation, built against the Model Context Protocol and agent-to-agent communication. 3rd Prize.

MCP
A2A
Multimodal
Autonomous Agents

3rd Prize — MCP × A2A Hackathon

Multi-Agent System for Technology & Market Evaluation
Agents & MCP

A multi-agent platform where specialized researcher, market-analyst, technical-evaluator and synthesizer agents collaborate on proposal evaluation, with a full-stack interface for running and reviewing results.

CrewAI
LangChain
GPT-4o
Flask
React
+1

Full-stack multi-agent decision support

Engineering Simulation Analysis Agent
Agents & MCP

An agentic workflow that parses simulation logs, detects anomalies and converts raw output into structured technical findings, with FAISS retrieval over specifications and prior analyses.

Python
CrewAI
LangChain
FAISS
PyTorch

Agentic anomaly detection over simulation logs

GPU & Local Inference Lab
Featured
GPU Systems

A systems-focused lab of six experiments on local LLM inference, GPU memory, scheduling and fine-tuning — measuring what most demos skip.

PyTorch
CUDA
Local LLMs
Quantization
QLoRA
+2

Six experiments: inference, memory, scheduling & fine-tuning

Edge-to-Cloud ML Product System
AI/ML

Python pipelines for real-time ingestion and TensorFlow Lite inference on connected hardware, paired with React dashboards for live telemetry and device health.

Python
TensorFlow Lite
React
FastAPI
AWS

Edge-to-cloud ML product under real hardware constraints

AI-Driven Financial News Reporter
NLP

Retrieval, clustering, ranking and summarization over continuously updated market sources.

FastAPI
PostgreSQL
pgvector
Gemini
Cohere

Retrieval + summarization over live market sources

Mass Communication Platform
Web Dev

A full-stack communication platform for contact management, campaign workflows and delivery-state tracking.

FastAPI
Next.js
MongoDB
REST APIs
Full-Stack

Contact management & delivery-state tracking

Human vs AI Writing Classification
NLP

An NLP classifier distinguishing human- from AI-written text, evaluated with precision/recall/F1 and documented reliability limits.

Python
NLP
Scikit-learn
Text Classification

Evaluated with precision/recall/F1 + documented limits

Bootstrap & Permutation Analysis
Data Science

Resampling workflows producing empirical uncertainty distributions, comparing CI-based and permutation-based conclusions.

Python
Bootstrapping
Permutation Testing
Statistical Inference

CI-based vs permutation-based inference compared

Graduate Admission Prediction
Data Science

A logistic regression model with coefficient interpretation, class-balance handling and stated bias limitations.

Python
Logistic Regression
Scikit-learn
Statistical Inference

Coefficient interpretation + stated bias limitations

Quick Lube & Smog Website
Web Dev

A production website designed and developed end-to-end for a local automotive business — services, trust signals, and contact, from requirements through deployment.

Next.js
TypeScript
Frontend
Deployment
SEO

Live production site, owned end-to-end

More experiments, research notebooks, and open-source work on GitHub.

View GitHub Profile