Every claims-automation pitch starts at the same place: the adjuster. Automate the intake, automate the triage, let the machine draft and the human decide. What the pitch usually skips is the data underneath — and in delegated authority business, that data is the bordereaux: the periodic reports of premiums, claims, and risks that flow from MGAs and coverholders to the insurers and Lloyd’s carriers whose capacity they write on. Bordereaux are famously late, inconsistent, and cleaned by hand. On November 2, 2026, Five Sigma and Amarillo Technologies announced a partnership that starts at that layer instead: AI-native claims management on one side, bordereaux and delegated authority data management on the other.
For brokers, this matters more than another claims-platform launch. If you place business into delegated authority programs — or your clients’ capacity sits in the London Market — the quality of bordereaux data decides how fast claims get seen, how cleanly a program’s performance reads at renewal, and whether the automation layered on top is working from facts or from a spreadsheet somebody fixed by eye. Here is what was announced, why the data layer is the real story, and the questions worth asking.
What was announced
Five Sigma, which describes itself as an AI-native claims management platform for P&C insurance, and Amarillo Technologies, a London-based insurtech specialising in delegated authority data management, announced a strategic partnership on November 2, 2026, aimed at helping insurers and MGAs modernise claims operations across the UK and London Market — the market whose capacity still turns on the Lloyd’s and January 1 renewal cycle we track every year.
Five Sigma’s side of the pairing is claims automation. Its platform supports the claims lifecycle from first notice of loss (FNOL) to settlement, and its Clive™ product is described as a multi-agent AI claims expert. The company says the platform lets insurers, MGAs, and reinsurers automate routine work by default, keep adjusters focused on decisions that need judgment, lower loss adjustment expense, and shorten cycle times. Those are the company’s claims for its platform, stated in its own release.
Amarillo’s side is the data. Its Fusion platform automates the collection, processing, and validation of bordereaux data. Insurers, MGAs, and brokers use it, per the release, to reduce manual effort, improve data quality, strengthen compliance, and see clearly into delegated authority operations.
The partnership itself, as announced, is a go-to-market pairing: joint go-to-market initiatives and shared industry networks, expanding market reach, and connecting each company’s clients with the other’s expertise. The release describes a partnership focus on the UK and London Market in particular. It does not announce a shipped technical integration, a joint product, pricing, or named customers — read it as two specialists agreeing to sell alongside each other, and judge it on that basis.
Why bordereaux is the real problem
The quotes in the release point at the data layer, not the adjuster. “Insurers and MGAs want to modernise claims, but fragmented data and manual processes keep holding them back,” said Michael Krikheli, CTO and co-founder of Five Sigma. “Delegated authority is where that friction shows up most not only in the London Market but across the world. Amarillo understands that problem deeply, and this partnership means our clients can get expert help on both sides of it.”
Zoe Steele, Managing Director of Amarillo Technologies, made the same point from the data side: “Our customers want better data quality and less manual effort across the whole delegated authority chain, including claims.”
That framing is worth taking seriously, because it matches how delegated authority actually breaks. A carrier gives an MGA or coverholder authority to write business on its paper. The carrier’s window into what was written, what premium came in, and what claims are developing is the bordereaux — often a spreadsheet, often monthly, often in whatever format the MGA’s system produces. Claims data rides in that flow. If the bordereaux is late, miscoded, or inconsistent from one month to the next, everyone downstream works from a partial picture: reserves set on stale information, program performance argued over instead of read, and renewal conversations — the kind that decide where placements go when a program closes — conducted on numbers neither side fully trusts.
Now layer claims automation on top of that. An AI claims platform fed by dirty bordereaux data does not fix the errors; it processes them faster and with more confidence. The pairing announced here is notable because it at least names the correct order of operations: clean, validated delegated authority data first — Amarillo’s stated job, automating collection, processing, and validation — and automation second. Whether the partnership delivers that order in practice is a different question, and it is the one brokers should press on.
What brokers and MGAs should ask next
You may never buy either platform. But if your programs, your capacity partners, or your clients’ claims run through delegated authority arrangements, the answers to these questions show up in your renewals:
1. Where does our bordereaux get cleaned today, and by whom? If the honest answer is a person reformatting spreadsheets before each submission, that is the manual effort both companies are pointing at — and it is the first thing to price against any automation claim.
2. Is claims data validated before it reaches the carrier, or reconciled after? Amarillo’s stated model is validation at processing time. Ask your MGA and carrier partners which side of that line their bordereaux sits on, because errors found after a claim is booked are the expensive kind.
3. If we automate claims handling on this data, what happens to the records that fail validation? Automation on the clean majority is easy to demo. The exceptions — the miscoded, the duplicated, the missing — are where delegated authority claims go wrong. Ask who sees them and how fast.
4. Is this a product integration or a referral? As announced, this partnership is go-to-market: shared networks and mutual expertise, not a launched joint product. That is not a criticism, but it changes what you are being sold. Ask what exists today versus what the partnership intends to build.
5. What can you show me from a live program? Cycle times, LAE, data-quality improvements — ask for them from an operating delegated authority book, not a platform demo. We have made this point about claims-speed claims before, in our piece on what to tell a client when the claims shop says it’s four times faster: the number worth believing is the one measured on real claims.
The honest limits
Three cautions, because this is a vendor announcement and should be read as one.
First, there are no results in this release. No named insurer, MGA, or broker is quoted using the pairing; no cycle-time, data-quality, or cost figures are offered for the partnership itself. The benefits language — modernise, improve efficiency, accelerate adoption — is the standard vocabulary of a launch announcement. The mechanism is plausible: validated bordereaux data is genuinely the weak layer in delegated authority, and both companies’ descriptions of their own platforms are at least aimed at real work. Plausible is where the evidence stops.
Second, a partnership announcement is not an integration. As released, this is two companies aligning go-to-market efforts and client introductions. Brokers evaluating either vendor should evaluate the platform in front of them, at the price quoted, on the timeline promised — and treat the partnership as a direction, not a deliverable.
Third, cleaner data does not settle the judgment questions. Bordereaux validation can make a claim visible sooner and coded correctly; it cannot decide coverage, set the reserve, or negotiate the settlement. Five Sigma’s own framing concedes this — adjusters kept “focused on decisions that need judgment.” Hold both vendors to that division: machines for the flow, people for the calls, and data clean enough that the people are deciding on facts.
The direction, though, is the right one. The London Market’s delegated authority machine runs on reported data, and the reported data has been the dirtiest part of the chain for as long as anyone has worked in it. A claims-AI vendor pairing with a bordereaux specialist is an admission, in public, that automation was never the first problem. The data was.
Informational only — not legal, coverage, or claims-handling advice. Partnership details and quotations are from the Five Sigma and Amarillo Technologies announcement of November 2, 2026; vendor claims are attributed as claims and have not been independently verified.
Sources: Five Sigma and Amarillo Technologies partner to advance insurance claims modernisation for London Market (EIN Presswire, Nov 2, 2026) · Five Sigma