AI and Machine Learning
Large language models (OpenAI, Anthropic, open-source via Ollama), retrieval-augmented generation (RAG), vector databases (Pinecone, Weaviate, Chroma), prompt engineering, model evaluation, and fine-tuning.
Technology
My tech stack is chosen for what ships, not what is trending. I focus on technologies that are reliable, well-documented, and proven in production.
Everything from model selection and retrieval pipelines to frontend frameworks and deployment infrastructure. Chosen for reliability, not hype.
What this includes
Every technology is vetted for production use - no black boxes, no vendor lock-in.
Large language models (OpenAI, Anthropic, open-source via Ollama), retrieval-augmented generation (RAG), vector databases (Pinecone, Weaviate, Chroma), prompt engineering, model evaluation, and fine-tuning.
PHP (Laravel), Python (FastAPI, Flask), JavaScript/TypeScript (React, Next.js), HTML/CSS, modern CSS frameworks. Fast to develop with, fast to serve.
Docker, nginx, Linux server administration, CI/CD pipelines, cloud hosting (AWS, DigitalOcean), managed databases, and monitoring.
Figma for design, modern CSS with custom properties, responsive design, accessibility (WCAG), performance optimization (Core Web Vitals).
Approach
No technology for technology sake. Every tool is chosen based on the specific problem, team constraints, and production requirements.
The best architecture is the simplest one that handles the requirements. Complexity is the enemy of reliability.
Every project is built with deployment, monitoring, and maintenance in mind from day one.
Clients always know exactly what technologies are in their stack and why.
Outcomes
Built with technologies that are well-supported, well-documented, and battle-tested in production.
Familiar toolchains mean faster development, fewer surprises, and quicker time to market.
Clean architecture, clear documentation, and technologies that your team can own long-term.
Let's work together