AI and machine learning
Computer vision, neural networks, model evaluation, and applied LLM/RAG workflows for product teams.
- PyTorch and TensorFlow
- YOLO-style vision systems
- RAG and agentic workflows
Available for AI engineering work
Production AI for product teams that need more than a demo.
I build computer vision products, LLM/RAG systems, and full-stack AI platforms that move cleanly from prototype to production.
- 99%+
- Vision accuracy
- 10+
- Projects built
- 3
- Delivery layers
LLM/RAG
FastAPI
Secure APIs
What I build
From model logic to a product people can use.
I pair applied AI with the API design, interface, and operating discipline needed to make it useful beyond the prototype stage.
Full-stack delivery
Production APIs, frontends, dashboards, and integrations that keep AI outputs usable and observable.
- FastAPI and REST APIs
- React interfaces
- PostgreSQL data layers
Deployment and security
Containerized services, authentication, CI/CD, and practical hardening for systems that need to run.
- Dockerized deployment
- JWT/OAuth patterns
- Cloud-ready automation
Selected work
Proof over decoration.
Computer Vision
Traffic Symbol Recognition
Real-time sign classification with deep CNN workflows and 99%+ reported accuracy across 43 classes.
View case study
Medical AI
Cardiovascular ML Ensemble
Ensemble modeling for cardiac-risk prediction using explainable, testable machine-learning workflows.
View case study
Full Stack
CRM Platform
Role-aware backend and dashboard architecture for customer data workflows and operational reporting.
View case studyCore stack
Tools that carry a build from sketch to scale.
The choices here support one workflow: train, integrate, ship, observe, and improve.
Python
PyTorch
OpenAI API
FastAPI
React
PostgreSQL
Docker
Security
Current focus
Engineering work that stays clear under pressure.
The focus is practical: make models explainable, interfaces understandable, and the handoff from experiment to production deliberate.
Model-to-product workflow
Translate trained models into APIs, dashboards, monitoring loops, and user-facing flows.
Useful LLM systems
Design RAG, retrieval, evaluation, and guardrail patterns that make model behavior easier to trust.
Frontend polish
Build interfaces where visual polish helps users understand what the system is doing.
Next step
Have an AI product to ship?
Bring the messy brief. I can help turn it into a prototype, production service, or polished user-facing experience.