AI for Business
Most AI projects stop at the demo. I build the infrastructure and the process that take them to production. I led DevOps for a GenAI platform used by ~60,000 employees.
What I do
- LLM platforms — Terraform modules, assistants, knowledge-source onboarding, training and fine-tuning pipelines, governance of sensitive data.
- AI agents with the permissions of a new admin — least privilege, an audit log, human approval for irreversible actions, and a kill switch.
- RAG as a data project — every source has an owner, freshness and access controls, and an evaluation set. Answers cite sources, not the model’s “memory”.
- FinOps for LLMs — cost per request and per token measured from day one, not after the first invoice.
How I work
Every deployment starts with “what should this change in a specific team’s work” and ends with verifying the result — not a “wow” demo. Have an AI prototype that needs to reach production? Get in touch.
Get in touch