What I Do
AI engineering and full-stack expertise
Production AI agents, full-stack apps, and the habits that keep them reliable after launch.
Core expertise
Agentic AI and senior engineering
AI agents come first. The rest cover the apps and the AWS infrastructure under them.
AI Systems
AI Engineering
AI agents shipped into real apps, with RAG that grounds each answer in the app's own data.
- Agent orchestration & MCP
- Eval suites & guardrails
- Token economics & FinOps
- On Amazon Bedrock, Azure or self-hosted
Agent Harness & Deployment
Safety rails that decide what a production agent can touch. Runs on managed APIs, self-hosted infrastructure, or fully air-gapped.
- Typed tool surfaces and MCP servers
- Sandboxed execution and permission models
- Eval gates and regression suites
- Tracing, audit logs, PDPL and DIFC alignment
Web & Cloud
Full-Stack Engineering
Web apps, APIs and backend services, built in the language the project needs.
- Python, TypeScript, Go and Java
- Next.js, React and React Native apps
- APIs, queues and background jobs
- CI/CD on GitHub Actions or Azure DevOps
AWS Infrastructure
Next.js on AWS with SST and OpenNext, billed per use. It is the stack this site runs on.
- Vercel/Firebase to AWS migration
- CloudFront, Lambda and DynamoDB
- SST v3 & OpenNext deployment
- IAM & CI/CD with OIDC
Search
Ranking and AI citations
Stack
Built in the language the job needs
Each project picks its own stack, from what the team already runs and what the problem needs. Each case study lists the stack it shipped with.
Languages
- Python
- TypeScript
- JavaScript
- Go
- Java
- SQL
- C#
- PHP
Frameworks
- FastAPI
- PyTorch
- Node.js
- Next.js
- React
- React Native
- Vue
- Spring Boot
- .NET
- Symfony
AI platforms
- Amazon Bedrock
- Microsoft Foundry (formerly Azure AI Foundry)
- LangGraph
- MCP servers
- Ollama and vLLM
Cloud and delivery
- AWS
- Microsoft Azure
- Google Cloud
- Azure DevOps
- GitHub Actions
- GitLab CI
- Docker
- Terraform