Acceptance-tested on your own documents · #1 on NetArena MALT (third-party-run)

Market-data and research platforms

Answers your subscribers
can check.

Generic AI features answer fluently, but computed figures — growth rates, ratios, cross-document comparisons — often arrive without a trace. A wrong number under your brand is a reputational and contractual problem. The API returns each answer with its page, its table and its calculation, rendered in your own interface.

Who it’s for

Product, data and content teams.

Terminals, research platforms and data products whose subscribers rely on the figures you publish — and the teams who answer for those figures.

Who buys

The people who own the product and the data

  • The CPO or head of AI product
  • The CTO
  • The head of content and data operations
  • Partnerships or corporate development

Who uses

Your subscribers and your data team

  • Your subscribers, through cited answers in your interface (Job 1)
  • Your data-operations staff, for collection, QA and “explain this number” (Jobs 2 and 3)

Jobs we do for you

Every figure with its page and its math.

Each answer arrives with the evidence your subscribers and your content team need to check it: the source page and table, the calculation, and any conflict between documents flagged.

Job 1 · For your subscribers

Cited answers embedded in your product

Documents
Your filings, transcripts, research and archive content.
What you get back
The full evidence behind every answer, rendered in your interface: the answer, its cited page and table, the calculation trace, and conflicts flagged.
Measured by
  • Accuracy on your gold question set.
  • Citation validity.
  • Wrong-number support tickets.
  • Response time.
  • Cost per answer.
Targets, measured on your documents

Job 2 · For data operations

Data collection and QA

Documents
Non-standardized, international and scanned historical filings.
What you get back
  • Data points with lineage.
  • Trace-to-source on samples of your existing database.
  • Conflicts between your database and the source, flagged.
Measured by
  • Cost per data point.
  • QA error rate.
  • Time to availability after filing.
  • New coverage.
Targets, measured on your documents

Job 3 · For both

“Explain this number”

Documents
Any figure already in your database.
What you get back
Its source page and the calculation steps behind it.
Why it matters
A trust feature for your subscribers, and a QA tool for your content team.

Which license fits

Platform license only, on Enterprise terms.

Answers shown to your own customers, or built into your product, are licensed under Enterprise platform terms.

Using answers only inside your own firm, not in a product? Find your desk or see both licenses.

Acceptance test

How we’d prove it: head to head, on your gold set.

Not a demo on our documents. A side-by-side run against the feature your subscribers use today, scored on the questions they actually ask.

Your gold set

Real subscriber questions

Questions from your own logs, with answers written by your content team as the answer key.

Head to head

Against your current feature

Accuracy, citation validity, conflicts flagged, response time and cost per answer, side by side.

Agreed first

Pass criteria in writing

Acceptance test on your own documents, agreed in writing before you sign. We report the results as measured, misses included.

How acceptance testing works

Questions we hear

What platform teams ask first.

“We’ll build it with our current AI provider.”

Run the head-to-head acceptance test first. Same gold set, same scoring, side by side.

“What about scale and service levels?”

They’re defined in your Enterprise order form.

“Our content licenses restrict third-party processing.”

Deploy inside your environment. An Enterprise platform license can run in your cloud account or on-premises, scoped with you.

“Exclusivity?”

Talk to us. Book 30 minutes.

How we prove it

Tested on your documents, not just ours.

Every license and engagement is measured on your own documents and on questions your team has already answered. We agree the pass criteria with you first — figures right within an agreed tolerance, every citation checked against its page, known conflicts flagged — and report the results as measured, misses included.

  • Pilots: Acceptance-tested on your own documents before you pay the balance. Fixed fee, capped.
  • Enterprise: Acceptance test on your own documents, agreed in writing before you sign.
  • Early Adopter: Your acceptance plan is agreed in week one, and six months of development support are aimed at passing it.

How acceptance testing works

A third-party-run result

#1NetArena MALT on AgentBeats
100% / 100%Correctness and safety over 2,500 queries
0.056 sFastest of the perfect runs

A network-policy benchmark, not a finance one. We show it because it’s how we work: define correct, test at volume, publish the run.

Standing checked October 6, 2026.

Talk to us

Discuss a platform license.

Tell us what your subscribers ask, where the content lives and how answers would appear in your product. We’ll come back with a draft head-to-head test plan and a scoped proposal.

We reply within one business day. Prefer a call? Book 30 minutes.

We use this only to respond to your request. See our Privacy Policy.