Read a whole body of documents
Reports or filings, read and sorted for what matters to you.
Research and monitoring that cites its sources
Some research is too big to do by hand: thousands of filings, reports or websites. We use AI to read it, a strict process to check it, and a source link on every finding, so a person can stand behind the result.
What we do
Watch companies and suppliers
Alerts when filings, officers or notices change, with the record attached.
Match records across registers
Link the same company across different sources, each match with a method and a confidence score.
Enrich your own records
Add verified company details to your CRM, with where each fact came from.
How we keep it honest
-
AI reads each item
-
A relevance check decides what matters
Tuned against a sample a person has graded.
-
Facts are extracted into a fixed structure
-
A second pass checks them
-
A person decides any ambiguous match
We never link records on a guess.
How we measure accuracy
Checked against a person’s sample
We measure the results against a sample a person has checked by hand.
Counts, not just percentages
We report the result as counts, not just percentages.
A source and a date on every finding
Every finding shows its source and its date.
No links on a guess
A person decides any ambiguous match. We never link records on a guess.
You choose where the AI runs
Tier 1
A hosted model, through a UK or EU endpoint
The best quality for the least effort. Before any of your data goes to a provider, we check that its business terms rule out training on it, and we tell you which region does the processing.
Tier 2
Inside your own cloud account
The model runs in your AWS or Azure account, under your access controls and your logs.
Tier 3
A private model on hardware in the UK
Open-weight models on your hardware or ours. Nothing goes to an outside AI provider. Our own research and document reading run this way.
Storage location and processing location aren’t the same thing. We’ll tell you both, in writing.
Data protection and good manners
- Where research touches personal data, we work out the lawful basis first, write it down and keep to it.
- Our research crawler, DatafaceResearch, identifies itself, reads robots.txt and backs off when asked. Read our crawler policy.
Related pages
Questions
Can you tell us when a supplier gets into trouble?
We can watch the public signals: late filings, insolvency notices, officer changes. You get an alert with the source. Judgement stays with you.
Where does the data come from?
Public registers and filings, public web pages, and your own records. Every finding shows its source and its date.
How accurate is it?
We measure it against a sample a person has checked, and report the result as counts, not just percentages.
Get in touch
Tell us what you need to keep watch on.
One paragraph is enough. We’ll reply with honest thoughts on whether we can help, and what it would cost.