Akur8 Discover Turns Competitor Rate Filings Into a Live Pricing Signal

Akur8's Discover platform streams competitor SERFF rate filings directly into the pricing actuary's workflow, cutting Branch's manual filing research from two and a half weeks to a few hours (Akur8, February 2026). It launches as commercial P&C premiums fell 1.2% in Q1 2026, ending a 33-quarter run of increases (CIAB, May 2026), a soft market where knowing the field's next move matters as much as GLM precision.

Tracking actuarial vendor platform acquisitions over the past 18 months, the bundling of pricing automation with competitive rate filing intelligence accelerated sharply across 2025 and 2026, and pricing actuaries at regional carriers we've spoken with describe a shift in what they worry about: not whether their indicated rate is defensible on cost, but whether it is timed correctly against a field that is already moving. Akur8, the transparent machine-learning pricing vendor used by more than 3,000 actuaries at 300-plus insurers across 40-plus countries (Akur8, January 2026), absorbed the rate-filing intelligence startup Matrisk on January 6, 2026, and folded its capability into a new module called Discover. Portage Mutual, a 140-year-old Canadian mutual serving eight provinces through more than 600 brokerages, selected Akur8's full platform in June 2026, with Akur8 citing Discover's filing intelligence as a distinguishing factor in the win (Fintech Global, June 23, 2026). The trade coverage has treated this as a growth story about a fast-scaling insurtech. It is also a story about what a pricing actuary is professionally attesting to when a rival's regulatory filing helped shape the number that actuary signs.

What Discover Actually Puts in Front of a Pricing Actuary

Discover ingests rate and rule filings from the System for Electronic Rate and Form Filing, the NAIC-run infrastructure that carriers use to submit filings in 47 states and that processes roughly 550,000 filing transactions a year (NAIC/SERFF, 2019 data). The platform refreshes its dataset on a rolling biweekly cycle and covers every P&C line and every state, spanning personal lines, small commercial, and specialty programs (Akur8 Discover product page, 2026). Where SERFF itself is a document repository, indexed by state and company but not built for cross-carrier analysis, Discover extracts structured data from the filing PDFs, including rate tables, rule changes, and territory and tier definitions, and layers dashboards on top: rate evolution by carrier and state, the largest market movers over a chosen window, and side-by-side comparisons of rating variables a competitor has newly introduced. A feature Akur8 calls Regulatory Radar goes a step further, surfacing the objections a state regulator raised against a similar filing in the past and the justification language that satisfied it, aimed at shortening the back-and-forth between an actuary's filing and a department's review. None of this replaces the loss-cost indication itself. It sits, in Akur8's framing, "upstream" of the modeling workflow, feeding market context into decisions the actuary was always going to make about segmentation, timing, and the size of a filed change.

Branch, an auto-focused insurtech that adopted Akur8's core platform in 2025 and expanded to Discover in February 2026, put a number on the time saved: a pricing team member reproduced two and a half weeks of coverage and filing research in a couple of hours (Akur8, February 2026). Samuel Falmagne, Akur8's CEO, framed the acquisition's purpose plainly: "Matrisk adds a critical new dimension to that mission. By bringing powerful filings search and competitive intelligence into our platform, we can now offer insurers the ability to understand their market context" (Akur8, January 6, 2026). That reframing, from pricing engine to market-context engine, is the part of the story that has not been examined for what it does to actuarial sign-off.

Why Timing Now Outweighs Precision in a Softening Market

For three years of hard-market casualty pricing, the actuarial problem was mostly about getting the cost estimate right: trend selection, loss development, the size of a social-inflation load. A carrier that filed a defensible increase rarely had to worry that the market would undercut it, because most of the market was also raising rates. That symmetry broke in the first quarter of 2026. The Council of Insurance Agents & Brokers' Commercial P&C Market Index recorded a 1.2% average premium decline across account sizes in Q1 2026, the first quarterly decline after 33 consecutive quarters of increases (CIAB, reported May 26, 2026). The softening is not uniform. Commercial property fell 5.5%, workers compensation fell 3.7%, and cyber liability fell 3.5%, while commercial auto continued rising 5.8% and umbrella rose 4.8% in the same quarter (CIAB, May 2026).

Line of business	Q1 2026 average premium change
Commercial property	-5.5%
Workers compensation	-3.7%
Cyber liability	-3.5%
D&O liability	-2.1%
Business interruption	-1.9%
General liability	+2.6%
Umbrella	+4.8%
Commercial auto	+5.8%

Source: Council of Insurance Agents & Brokers, Commercial P&C Market Index, reported May 26, 2026.

That divergence is exactly the environment in which filing intelligence stops being a nice-to-have and starts changing decisions. A carrier holding a positive property indication while the CIAB index shows the field averaging down 5.5% faces an adverse-selection problem that no amount of GLM refinement solves: raise into a softening line and the book skews toward insureds who cannot get a better price elsewhere. A carrier writing commercial auto, where the field is still moving up 5.8%, faces close to the opposite problem, with room to file the full indicated increase without conceding share. Filing intelligence does not change either indication. It changes whether the actuary recommends filing the full number, holding flat while re-underwriting the book, or pushing the increase into finer segments where competitors have not yet moved. That is a real information gain over the status quo, where a regional carrier without a market-intelligence function found out a competitor had cut rates only when renewal retention data confirmed it, months after the filing had already taken effect.

What Is the Actuary Certifying?

Property-casualty rate filings are not just spreadsheets; in most states they carry an actuarial memorandum in which a qualified actuary supports the filed rates as reasonable, not excessive, not inadequate, and not unfairly discriminatory, the four criteria the Casualty Actuarial Society has used to define a proper rate estimate since its 1988 Statement of Principles Regarding Property and Casualty Insurance Ratemaking, reinstated by the CAS board in 2021 for continued reference in U.S. regulated ratemaking. The certification is a statement about the relationship between the filed rate and expected future costs. It has never had to account for a second input: a live feed showing what every other carrier in the state just filed.

That second input creates a documentation gap Discover did not invent but makes far more consequential. An actuary's workpapers can show clean lines from loss data to trend selection to indicated rate change. What they typically do not show is whether, or how much, the final selected rate change (as opposed to the pure cost-based indication) moved because a competitor filed lower. If a carrier's indication supports a 9% increase and the actuary selects 4% because Discover flagged that the top five competitors in that state filed increases averaging 3%, the actuary has made a legitimate business decision, but the actuarial memorandum rarely distinguishes indicated-cost input from competitive-positioning input in a way a regulator, or a later auditor, could reconstruct. The Code of Professional Conduct that governs U.S. actuaries requires that professional judgment be exercised with objectivity and be free from bias introduced by the interests of any party. Competitive intelligence is not itself a conflict of interest. But when it starts calibrating the number an actuary is professionally attesting to, the honest answer to "what is being certified" gets harder to write down, and no state rate-filing statute has been updated to ask for that distinction explicitly.

The Herding Risk in a Shared Signal

A second-order effect follows from adoption scale. Akur8 added more than 50 new clients globally in 2025, including more than 25 in North America, on top of annual recurring revenue growth exceeding 50% (Fintech Global, March 26, 2026). If a meaningful share of a state's active filers subscribe to the same filing-intelligence layer, that layer stops being a private edge and becomes a shared reference point the whole market prices against. That is a different failure mode than any one carrier underpricing. In a rising market, correlated caution among filers who all see the same competitor data would slow the market's ascent. In a falling market, the same correlation accelerates the descent: every carrier watching the same feed sees the same negative filings and reduces its own selected increase in response, which produces the next round of negative filings the following carrier's dashboard will show. Historical soft markets, including the workers' compensation trough of the early 2000s, took years to correct partly because individual carriers lacked visibility into how widespread the underpricing was until reserve deficiencies surfaced. A real-time shared filing signal could shorten that discovery lag. It could also make the descent faster and deeper before anyone breaks from the pack, because the tool that reveals the trend to everyone simultaneously removes the natural staggering that used to slow herd behavior down.

The Small-Carrier Equalizer

The same mechanism cuts the other way for competitive structure. Building an internal market-intelligence function that reads SERFF filings across 47 states has historically required either a dedicated competitive-intelligence team or a subscription to a research vendor, both of which favor carriers with the scale to afford them. Branch's two-and-a-half-week research cycle collapsing to hours is not just a productivity story; it describes work that a two- or three-person pricing team at a regional mutual could not previously do at all within a normal filing timeline. Portage Mutual's decision to adopt Akur8, including Discover, for a personal-lines book spread across eight provinces fits that pattern: a carrier with deep regional relationships but limited market-intelligence infrastructure can now see, filing by filing, what national writers are doing in its territories in near-real time, narrowing a data moat that scale used to guarantee the largest ten writers in a state. Whether that narrows pricing sophistication gaps or simply narrows how long the largest carriers can act ahead of the field before the rest catch up is the open question the next few renewal cycles will answer.

Integrating Filing Intelligence Without Losing Independence

Pricing actuaries who adopt a filing-intelligence layer do not need to wait for a state to update its rate-filing statute to protect the actuarial judgment the certification depends on. Four practices close most of the documentation gap now. First, keep the cost-based indication and the competitively adjusted selection as two distinct, separately labeled numbers in the workpaper file, so a reviewer can see the size of the gap between them and the stated reason for it. Second, log which specific competitor filings, and Discover citations, informed a selection change, the same way an ISO loss-cost citation would be logged, rather than referencing "market conditions" generically. Third, set an internal tolerance band, agreed with the pricing or actuarial committee in advance, within which competitive intelligence may influence the selected rate without additional sign-off, and require committee review for adjustments outside that band. Fourth, where a state's actuarial memorandum format allows narrative comment, use it: a short statement that the selected rate reflects both the indicated cost need and observed market positioning, with the indicated number stated plainly, gives a regulator the information to evaluate both inputs rather than inferring one from the other. None of these steps slow down the speed advantage Discover provides. They preserve the audit trail that makes the actuary's signature mean what it is supposed to mean.

Why This Matters for the 2026-2027 Filing Cycle

The commercial P&C market's divergence, property and cyber loosening while auto and umbrella keep hardening, means the next 12 months of rate filings will not move in one direction. Pricing actuaries who can see the field's filings in near-real time will make sharper calls about which lines to hold, which to chase down, and which to defend through segmentation rather than an across-the-board number. That is a genuine improvement in decision quality. It arrives, though, at the exact moment state insurance departments are also scrutinizing AI use in rate-setting more closely, and an actuarial memorandum that cannot show how competitive intelligence and cost-based indication combined into a final selected rate is a weaker document to defend in that environment than the same filing was three years ago. Carriers adopting Discover, or any of the filing-intelligence tools likely to follow it, should treat the workpaper discipline described above as part of the rollout, not an afterthought bolted on after a regulator asks the question first.