Named by AI is not the same as cited by AI.
I matched 102 company domains against captured AI answers and their source entries. The report keeps agent visibility separate from evidence ownership.
Public data. Clear decisions.
I combine databases, spreadsheets, formulas and LLM evidence to answer awkward commercial questions. You can inspect the method, rows and limits behind every result.
Working stack
The public reports show the output. This is the wider toolkit I bring when the source lives in databases, spreadsheets or AI platforms.
Use joins, filters and grouped measures to turn operational tables into a clean analytical base. Keep the query logic visible and repeatable.
Latest builds
I matched 102 company domains against captured AI answers and their source entries. The report keeps agent visibility separate from evidence ownership.
I joined Land Registry prices, ONS earnings and Bank Rate data. The model compares nine English regions and states where the proxy stops being reliable.
Research ledger
Every item links to the visible result. The dashboards also expose their source rows and method.
I will help you find the useful question, build the model and explain what the evidence actually supports.
Book a data review