Helping tech founders build AI applications that actually scale.
AI Augmented Engineer specializing in Python, Javascript and seamless AI Integrations.
Core technologies: Next.js, Nest.js, React, Python, Django, TypeScript, PostgreSQL, AWS, Terraform, Docker, OpenAI, Supabase, Stripe, Tailwind.
Selected work
Things I've built and shipped.
From greenfield AI products to SOC 2-ready cloud foundations — eleven platforms, most built solo from the first commit.
AI & LLM
OGE — Operational Gaming Environment
Built from the first commit — multi-provider LLM advisors, production RAG, WebSocket multiplayer.
Lead full-stack engineer
DevOps & Infra
Dayonai
Delivered a SOC 2-ready AWS foundation from scratch for a 20-engineer org, entirely as Terraform IaC.
Sole DevOps engineer
AI & LLM
Ellavator AI (Ella)
Multi-tenant LLM product with fine-grained access control, migrated OpenAI → OpenRouter in production.
Full-stack engineer
Full-Stack SaaS
HeroGeneration
Built a HIPAA-compliant caregiving platform end to end, with real-time collaboration and subscription billing.
Sole developer
Full-Stack SaaS
Quulio
Architected an all-in-one business and finance platform for self-employed professionals in Finland.
Founding full-stack developer
AI & LLM
Leo
Built an AI transcription and research platform for historical documents from scratch, including credit-based billing.
Sole developer
How I work
Four things I don't compromise on.
Ship to production, not to demo.
Every project listed shipped with real users, real billing, and real monitoring.
AI that's debuggable.
Retrieval and LLM systems built with evaluation, tracing, and fallbacks — not prompt-and-pray.
Own the whole stack.
Schema through deployment. Most of these were built solo, from the first commit.
Built to be audited.
Infrastructure as code, least-privilege access, environments that separate cleanly.
- 4+
- years shipping
- 11
- production platforms
- 5
- solo-built end to end
- SOC 2
- ready infra delivered
Engineering notes
Things I've learned the hard way.
· 6 min
We spent three weeks blaming the prompt. It was the retriever the whole time.
On debugging RAG in production, and why the failure is almost never where you think.
· 4 min
7 things I do differently after 4 years of shipping.
Engineering lessons that only show up after the project has been live for a while.
Have something you want built?
Tell me what you're working on. If I'm the right fit I'll say so — and if I'm not, I'll tell you that too.