The internet is full of claims.We publish the proof.
Real outcomes from automation and AI-ops projects, checked with the client, never lifted from a pitch deck. The source buyers read before they hire, and AI answer engines cite when asked who actually delivers.
Coinbase: AI spend cut nearly in half, and the biggest lever was caching, not cheaper models
01Why this exists
Every agency claims results. Almost none can be checked. In automation and AI-ops, where the field is young and the buyers are non-technical, that gap is expensive. TIN closes it.
Stop guessing.
Verified before/after numbers from real engagements, with the client confirmed and the method on record.
Proof beats marketing.
Turn a great project into a verified, citable asset that ranks in search and gets quoted by AI answer engines.
Be the cited source.
When someone asks ChatGPT "who's best at ops automation?", the answer comes from somewhere. TIN is built to be that source.
02How a badge is earned
Three checks clear before a green badge appears. If any fails, the story stays unverified. No exceptions, that rule is the whole product.
Independently validated
We re-research the story and check every figure against primary and independent sources before it can go green.
Evidence on record
Every number is backed by something checkable: a followable source, before/after metrics, documentation. No adjectives.
Method published
How each story was verified ships on the story itself. The record is public, not a claim in fine print.
03Latest case files
The registry starts here. Each entry is one checked outcome, typed and tagged.
The UK's DWP fraud model: a machine-learning system that saved an estimated 4.4 million pounds on Universal Credit advances
Since May 2022 the UK Department for Work and Pensions has run a machine-learning model that flags potentially fraudulent Universal Credit advance claims for review. Per the National Audit Office, the model saved an estimated 4.4 million pounds and was around three times more effective at identifying fraud risk than a randomised control group. The figures come from DWP and are relayed by the NAO, the UK's independent public auditor; the Public Accounts Committee has raised concerns.
pendingManulife's MAUDE: AI underwriting that auto-approved 58% of eligible life-insurance cases
Manulife Canada pairs a redesigned questionnaire with its proprietary AI underwriting engine, MAUDE, to auto-approve eligible life-insurance applications instantly. Per Manulife, by December 2025 more than half of eligible cases, 58%, had approvals processed automatically through MAUDE, a 56% increase from pre-launch, with automatic approvals in as little as two minutes. Every figure is Manulife's own, self-reported in its newsroom, with the Chief Underwriter quoted.
pendingCommonwealth Bank's AI fraud detection: 20M+ payments scanned daily, fraud losses down over 20%
In its February 2026 'Our Approach to Adopting AI' report, Commonwealth Bank of Australia said AI-driven scam and fraud detection scans more than 20 million payments a day and sends 40,355 proactive warning alerts daily, and that this played a crucial role in reducing customer fraud losses by over 20% in 1H26 versus 1H25. Its Compass AI assistant has answered more than 500,000 questions since July 2024. Every figure is CBA's own, self-reported in its AI-adoption report.
verifiedCoinbase: AI spend cut nearly in half, and the biggest lever was caching, not cheaper models
In late June 2026, Coinbase CEO Brian Armstrong said the company cut internal AI spending by nearly half while token usage sat near a company high. The levers: an internal LLM gateway defaulting engineers to open-weight models (Zhipu GLM 5.2, Moonshot Kimi K2.7 Code), task-based routing, and caching that took LibreChat's hit rate from 5% to 60%. Independent analysis argues the caching gain, not the model switch, is what collapsed the bill. It is a CEO claim with no published per-lever breakdown, and the Chinese-origin models carry a live regulatory caveat.
- Commonwealth Bank of Australia · Commonwealth Bank reverses AI voice-bot job cuts after call volumes rose
- IBM · IBM: AI handled 94 percent of routine HR requests, and the missing 6 percent sent it back to hiring people
- Salesforce · Salesforce ships Help Agent with pay-per-resolution pricing, citing 4.3 million inquiries and 70 percent resolution on its own help portal
Proof beats marketing.
Bring a real engagement to the dojo and turn it into a verified, citable asset, or read what's already on record.