Ilya Shkura
Numbers that hold up, checked against the data they came from.
I build tools that check figures against raw data, and I use them for clients: sales reports, checks of reports someone else wrote, reconciliation of files that should agree. This site collects the work, the projects and the cases.
Now
Freelance
Sales data, checked to the cent
Sales reports from your export, checks of reports you already have, and reconciliation of sales against payouts. Rules agreed in writing before counting.
Services and cases →Project
Iterum
A system that checks every figure of an analytical report against the data it came from, and the public evidence of what its checks stop and what they miss.
About the project →Selected cases
All cases →
€285,611between Shopify sales and payouts, explained line by line
8 of 35recounted figures wrong in an AI-written sales report
36/36figures of a quarterly report matched an independent recount
37.2%of a fund platform's value in suspended or delisted securities
Self-initiated case studies on synthetic data in real export formats.