How we work together.
Three ways to collaborate. Short list of clients at a time. Contract work.
Data architecture
Canonical models, lakehouses, governance. The foundations that hold up for years.
- Master data management
- Entity resolution
- Data lineage & governance
Analytics engineering
Reproducible pipelines, dbt, orchestration. From chaos to a warehouse you can trust.
- ELT with dbt + Airflow
- Testing and documentation
- Data observability
AI products
Agents, RAG, classification. AI that actually fits into business workflows.
- LLM agents with traceability
- RAG over internal data
- Evaluation and monitoring
Three phases.
We start with a thirty-minute call. If it fits, I send a short proposal — scope, price, milestones, end date. I work on contract, not open retainer.
I work in two-week sprints. Demo every fifteen days with real data, not slides. You decide what goes into the next sprint.
When closing, I leave the project audited: documentation, runbooks, transferred access. After that I'm available for questions, but the project lives without me.
Do your data look like a pile of spreadsheets?
The first call is free, thirty minutes. I listen to the context, tell you whether I can help, and if not, refer you to someone who can.