Service
AI Agents & LLM Systems
LLM features that hold up once real data and real users arrive: tool-using agents, safety passes, and assistants scoped to your own data.
What I deliver
LLM features that hold up once real data and real users arrive: tool-using agents, generation pipelines with a safety pass in front of them, and assistants that answer over your own data without being able to reach anyone else's.
Technologies
Claude and Gemini through their own APIs or through Vertex AI, the Model Context Protocol for tool use, .NET and Python for the services around them, and Postgres or BigQuery underneath. Retrieval where it earns its place, and a deterministic lookup where that is genuinely the better answer.
What to expect
The boundaries first and the clever part second. What the model may read, what happens when it gets something wrong, what a question costs, and a log that shows afterwards what it actually did. Most demos fall over on exactly those four things.
Case studies
Marketing Data Platform & Reporting Dashboard
A BigQuery warehouse and reporting dashboard for a US marketing agency, replacing a spreadsheet workflow with one source of truth per client, plus an assistant that answers questions over it in plain language.
Smart Marketing Agent: Multi-Brand Engine
An AI-assisted marketing agent that generates and manages brand-safe campaign content, architected multi-brand from day one so it can scale across brands without a costly rebuild.
Hrolbot
Multi-agent AI platform running on a Proxmox VM. Autonomous Claude Code workers coordinated through an MCP mesh for code review, builds, testing, and system operations.
Contact
Tell me what you need
A couple of lines is enough. I reply within a day.
Bigger project? Use the full project form, with budget and timeline