Services
Three things that usually follow from one another in practice: first get the data right, then build on it.
checks that run along
Data warehousing
A data warehouse that holds up, even when the source changes. I build the modelling, the load processes and the checks that belong underneath, so every figure in a report can be traced back to its source. Usually on SQL Server or Azure, in existing environments with existing agreements.
- SQL Server
- Azure Data Factory
- T-SQL
- Azure SQL
checks that run along
AI software on your own data
AI grounded in your own data. That means searching your documents with source references, or asking questions about your figures where the answer comes from the data warehouse rather than from a language model that guesses. The real work is in the boundaries: which data may a model see, and how do you verify the answer.
- Python
- FastAPI
- Azure OpenAI
- Azure AI Search
- RAG
Applications on your data
A report is not always the answer. Sometimes it is an application where people enter, review or record data. I build those applications on top of the same data layer, with authentication through the account your organisation already uses.
- Next.js
- TypeScript
- React
- Entra ID
Does this match what you need?
Describe the situation and the question in a few lines. I will tell you whether I am the right person, including when the answer is no.