Building AI that actually works in the real world
This is where I experiment, build, and break things. My repositories range from research prototypes to production-ready systems, with a focus on AI/ML applications.
Currently obsessed with:
- Meta-learning and adaptive systems
- RAG architectures that actually work at scale
- Making AI more accessible through better tooling
- Edge computing and model optimization
AVM-CORE - Adaptive Validation Model w/ COmpositional REasoning (avm-core)
MATRIX - Thompson sampling meets reinforcement learning in high-dimensional spaces
Time Machine - Code evolution analysis system
DeepFake Detection - Computer vision system for detecting AI-generated content using advanced models
Drone Detection CV - Computer vision system for real-time drone detection and classification
EECS 194: SQL Mastery Learning - Interactive visualizations for database concepts (used in Berkeley courses)
ChatCHW - RAG-powered chatbot system for community health applications
I believe in building things that actually work. Most of my projects start as solutions to problems I'm facing in research or real applications, then evolve into more generalizable systems.
Languages I reach for: Python for ML/research, JavaScript for quick prototypes, C++ when performance matters, SQL for everything data
Favorite stack right now: Python + PyTorch + FastAPI + React + PostgreSQL
- Fine-tuning small language models for domain-specific tasks
- Building better vector databases from scratch
- Exploring multi-agent systems for collaborative reasoning
- Optimizing models for edge deployment
I'm always down to work on interesting problems, especially if they involve:
- Novel ML architectures
- Real-world AI deployments
- Educational technology
- Healthcare applications
Drop me a line if you want to build something cool together.
Most of my best ideas come from trying to solve problems I actually have.