03 — What I do
From the first idea to a product people use.
01
AI and agents
Agents that use tools safely, MCP servers and connectors, agent memory, guardrails and human approval.
02
Machine learning
Training and evaluating classifiers, sensitive-data detection, computer vision, voice generation on my own GPU server.
03
Backend and cloud
APIs, databases, queues, auth and permissions, deployment, monitoring, CI/CD.
04
Full-stack products
Web, mobile and offline-first apps, taken all the way to a live product.
05
Automation
Publishing, moderation, email and Slack flows with n8n, GitHub Actions and AI.
06
Video with code
3D characters in Three.js, AI voices in English and French, automated editing.
- Languages
- TypeScript, JavaScript, Python, Go, Rust, SQL. Also Java, C, C++, Dart.
- Backend
- Node.js, NestJS, REST, gRPC, WebSockets, microservices, Redis Streams, webhooks, MCP.
- Frontend & mobile
- React, Next.js, React Native, Flutter, PWAs, Three.js.
- Cloud
- AWS (EC2, ECS, S3, DynamoDB, RDS/Aurora, IAM, CloudWatch, GPU instances), Terraform. Cloudflare Workers, R2, D1, Access.
- Data
- PostgreSQL, SQL Server, SQLite/D1, MongoDB, DynamoDB, Redis, RocksDB, Supabase, Firebase.
- AI & ML
- PyTorch, TensorFlow, LLM APIs, agents, MCP, classifier evaluation, computer vision, text-to-speech.