Teardown

Insurance software · Deep dive

Comulate

AI accounting automation for large insurance brokerages — carrier statement ingestion, direct-bill reconciliation and commission recovery — that hit 8-figure revenue in under three years, then collided head-on with Applied Systems in court.

emerging

The question that decides it: Comulate's product only works if it can read and write policy data inside agency management systems it does not control — and Applied Systems, which Comulate's own complaint pegs at 80%+ of the enterprise-broker AMS market plus ownership of the Ivans data rails, has cut off its Epic access, won a preliminary injunction barring sales of Epic-derived product to new customers (Feb 2026), and watched Comulate shed customers and 20% of its staff. Does the Sherman Act case pry that access back open before attrition hollows out the 8-figure revenue base — or has Applied proven that any independent automation layer on Epic exists only at Applied's pleasure?

My take

HQ
San Francisco, CA
Founded
2022
Ownership
VC-backed (Series B; Feb 2025)
Funding
~$25M raised (TechCrunch / CB Insights, Feb 2025)
Valuation
Undisclosed
Revenue
8-figure, 'tens of millions' and tripled in 2024 (company, Feb 2025); under pressure from customer defections in 2026
Headcount
Undisclosed; company counsel said ~20% of staff was laid off by May 2026 amid the Applied litigation
Screen
Early breakout — founded 2022, raised $8M+ (~$25M total)
Published
2026-08-02
Web
www.comulate.com
Elsewhere
LinkedIn · Crunchbase

Founders and leadership

  • Jordan Katz Co-founder & CEO

    University of Michigan Ross grad and former product manager at Asana and Pathlight. With Mattheakis, a longtime friend from undergrad, he first tried to build revenue-management software for SaaS companies, concluded Silicon Valley did not need more software for itself, and pivoted to insurance broker accounting in 2022 after family and friends in the industry described the chaos of commission reconciliation.

  • Michael Mattheakis Co-founder & CTO

    Engineer with summer stints at Google and Facebook who joined Brex as roughly its 10th hire and spent nearly four years helping scale its platform and engineering team. Owns the statement-parsing ML, the AMS read/write integrations and the four issued US patents behind Comulate's reconciliation engine.

Snapshot

Comulate sells accounting automation to the largest US insurance brokerages: software that ingests carrier commission statements in any format, reconciles them to policies inside the brokerage’s agency management system (AMS), posts to the ledger in one click, and flags commissions carriers failed to pay. Founded in San Francisco in 2022 by Jordan Katz (ex-Asana) and Michael Mattheakis (ex-Brex), it reached 8-figure revenue in under three years, tripled revenue in 2024, skipped a Series A, and raised a $20M Series B co-led by BOND and Workday Ventures in February 2025, with customers including Lockton, Risk Strategies, IMA Financial and The Baldwin Group. Then the story inverted: in November 2025, Applied Systems — the AMS incumbent whose Epic platform Comulate depends on, and which had tried and failed to acquire it — sued for trade-secret theft, cut off its access, and by February 2026 held an injunction barring sales of Epic-derived product to new customers. Comulate answered with a federal antitrust suit. This is platform risk becoming a docket number.

Founding story

Katz and Mattheakis are longtime friends from their undergraduate days — Katz went from Michigan’s Ross School of Business into product management at Asana and Pathlight; Mattheakis did summers at Google and Facebook, then joined Brex as roughly its tenth hire and spent almost four years scaling its engineering team (The Org; Pulse2, Feb 2025). Their first idea, in Katz’s phrase, was “SaaS for SaaS” — better revenue-management software for their former employers’ peers. They dropped it as crowded and only incrementally improvable; Katz told TechCrunch (Feb 2025) that Silicon Valley did not need more software for itself.

The pivot came from outside tech: family and friends in insurance kept asking whether revenue software could be built for brokerages. Katz admits he barely knew what a premium was. Cold-emailing large brokers, they found back offices reconciling thousands of carrier statements by hand — a workflow ignored by AMS vendors and ERP players alike. Pilots converted fast enough that the founders incorporated as Ardent Labs, doing business as Comulate, in 2022, hiring from Brex, Asana, Plaid, Applied Intuition and Coalition. It is the inverse of the operator-founder story: zero domain pedigree, imported product velocity — which cuts both ways, as the fight over how they learned Epic would show.

How it works

The raw material is the carrier statement. A large brokerage receives thousands of commission statements a month from hundreds of carriers — industry sources count 13,000+ unique formats, roughly three-quarters PDFs never designed to be machine-read (ReSource Pro, 2025) — arriving as spreadsheets, PDFs, email and literal snail mail. Comulate’s ML extraction layer captures the transaction detail from all of them: policy numbers, insureds, premium, commission rates, PEPM figures, employee counts.

The hard part is not reading the statement; it is matching it. Comulate reconciles each extracted transaction to policy records inside one or more integrated AMS databases — Applied Epic, Vertafore BenefitPoint and AMS360, plus Microsoft Dynamics and Salesforce — even when carrier data does not match the brokerage’s (transposed policy numbers, renamed insureds, mid-term changes). Matched transactions post to the AMS or ERP ledger in one click, with write-back that corrects stale policy records — Heffernan Insurance Brokers reported 883 policies auto-corrected in one month of data (Rough Notes, Jan 2024) — while unmatched residue lands in an exception queue. On top sit carrier payables automation for the agency-bill side — flagging transactions against receivables, catching billing errors before notices of cancellation go out, routing discrepancies — and revenue intelligence: forecasting when commission revenue should arrive, surfacing variances, and drilling to policy level to recover underpaid commissions. Four issued US patents cover the machinery (company site, 2026). One mechanical detail now matters enormously: to build and demonstrate the Epic integration, Comulate operated a fake insurance agency, “PBC,” inside Applied’s system — a sandbox by Comulate’s description, fraudulent access by Applied’s, and the fulcrum of the injunction now constraining the company.

Product and business overview

Three named lines. Direct Bill Automation is the core: end-to-end statement ingestion, reconciliation and posting, sold on eliminating 90%+ of manual accounting work. Carrier Payables Automation, billed as an industry first, extends the engine to agency bill — the brokerage collecting premium and remitting to carriers — automating reconciliation and discrepancy management on the trust-account side. Revenue Intelligence is the analytics layer and expansion story: capturing every compensation detail on every statement lets it forecast revenue timing, track variances and guide recovery of missing commissions — a claimed 4% revenue lift (site, 2026). The go-to-market is unapologetically top-heavy: the stated ambition is the top 100 US brokers, and the customer list — Lockton, Risk Strategies, OneDigital, IMA, Baldwin, Hilb, INSURICA, Hylant, Trucordia, Inszone, Heffernan — is a roll call of national and PE-consolidated platforms. A February 2025 Workday Partner Program deal, the logic behind Workday Ventures co-leading the Series B, points at deeper ERP integration.

Business model and pricing

Enterprise SaaS, quote-based, no published price list anywhere in the public record — itself informative: this is negotiated, high-ACV software sold to CFOs. The ROI math is labor and leakage: agencies spend hundreds to thousands of hours monthly on statement processing (ReSource Pro, 2025), and Heffernan reported manual work down 70-90% within weeks of a five-week implementation (Rough Notes, Jan 2024). The recovered-commission claim lets the product price against found money rather than saved time. Margins are undisclosed; the cost base is ML inference, integration maintenance — and, since late 2025, litigation counsel.

Traction over time

MarkerJan 2024Feb 2025May 2026
RevenueUndisclosed; “dozens of large agencies” (Rough Notes)8 figures in under 3 years; tripled in 2024; “tens of millions” (company/TechCrunch)Undisclosed; customer defections reported in court
Customer-hours saved (company claim)260,000 by 2024 (TechCrunch); 350,000 trailing-12-month (PR)
Funding, cumulative~$5M~$25M~$25M (no new round disclosed)
HeadcountSmall ex-Brex/Asana/Plaid teamScaling post-Series BReduced ~20% (company counsel, in court)

Every pre-2026 number is company-supplied and unaudited; the multiples come off undisclosed bases. But the shape was real enough that Applied tried to buy the company in 2023 — and, per Comulate’s filings, again in 2024 and 2025. The 2026 column is the one to watch: by the May 2026 hearings, Comulate’s own attorneys told the court more customers had left and roughly 20% of the workforce had been laid off (Insurance Journal, May 26, 2026). A GetLatka estimate of $7M ARR circulating online is inconsistent with TechCrunch’s reporting; treat it as noise.

Market analysis

The US insurance brokerage industry generated roughly $140B of revenue in 2025 (Mordor Intelligence), and its back office runs on manual reconciliation. The direct software TAM — commission accounting for the few hundred brokerages that pay enterprise prices — is plausibly single-digit billions, and the structural forces point up: PE roll-ups (Hilb, Inszone, Trucordia and dozens more) keep gluing together agencies with incompatible books, multiplying reconciliation complexity with each acquisition, while LLM-era document extraction has made the core technical problem tractable for the first time. The countervailing structure is the choke point: virtually every target customer runs Applied Epic or a Vertafore AMS as its system of record. Comulate’s antitrust complaint asserts Epic holds more than 80% of enterprise brokers, with Applied also owning Ivans, the carrier-connectivity rails (Insurance Journal, Jan 2026). Where the data lives inside two gatekeepers, the gatekeepers set the terms of automation — the thesis Applied is stress-testing in federal court.

Competitive intel

See the competitor table. Three tiers. At the top, the AMS incumbents: Applied (Epic, Ivans, the native Applied Recon product, and — per Comulate’s complaint — a campaign to steer customers to a preferred third party its own executive privately called uncompetitive) and Vertafore (Roper-owned, with its own commission modules). They own the policy data Comulate must match against; their weapon is not feature parity but access control, and Applied has used it. In the middle, Ascend, the best-capitalized independent, which launched a rival Direct Bill Automation product in September 2024 and distributes through ReSource Pro — the BPO whose offshore statement-entry labor is the actual incumbent in most accounting departments. Below, vBots, Synatic CommRecon, Fintary, Eventual and in-house RPA commoditize ingestion from beneath. Comulate’s defensible ground was never parsing — it was AMS write-back depth and the enterprise trust of the top 100. Both are exactly what the injunction impairs.

History and evolution

What people say

The case for. There is no G2 or Capterra presence — reviews live in trade press and company case studies, so weight them accordingly. The recurring theme is disbelief converting to enthusiasm. Heffernan’s CIO Kate Grasman told Rough Notes (Jan 2024) she did not believe the implementation promises until a five-week, two-IT-hours deployment delivered 70% less manual reconciliation and cut offshore outsourcing to zero. INSURICA’s chief innovation officer says the accounting team “raves about it with unique passion” (company site, 2025-26); Bond’s Jay Simons framed the thesis as “sleepy, but critical” (TechCrunch, Feb 2025). Existing customers, notably, are exempt from the injunction and have incentive to stay.

The complaints. The criticism is not about the software; it is about conduct and viability. A federal judge found Comulate likely breached its Applied contract by running a fake agency to demo and improve its product and by giving Epic access to unauthorized users — and, denying Comulate relief, attributed the damage to the founders’ “own misdeeds” (Feb-May 2026). Applied publicly accuses it of having “lied, cheated, and stole.” Customers have defected under legal uncertainty and, Comulate alleges, under direct Applied pressure. The revenue narrative rests entirely on unaudited company multiples. The uncomfortable version: Comulate proved the demand and the product — and may have handed its most dangerous competitor the litigation record with which to strangle it.

Outlook: the open question

Everything reduces to whether Comulate can restore legitimate, durable access to the systems of record. The bull case requires three things: the antitrust suit, or the discovery it generates, forces Applied into a settlement or court-ordered access regime reopening Epic to new-customer sales; the exempt existing base — Lockton, Risk Strategies, IMA and the rest — holds, keeping 8-figure revenue intact; and Vertafore, watching Applied’s legal exposure, chooses partnership over imitation, giving Comulate a second rail. The bear case requires less: the injunction stands through trial, new sales stay frozen, Ascend and ReSource Pro absorb the pipeline, attrition compounds, and the Series B becomes the last round — with Applied or a rival buying the assets at a fraction of the refused offers. The February 2026 ruling contained the tell for both sides: no derivative work, no reverse engineering — the product is Comulate’s own — but the access underneath it was likely obtained in breach. A company can survive being sued by its platform; it is much harder to survive being right about the monopoly and wrong about the contract. Watch the Illinois antitrust docket, any disclosed logo losses, and whether a Vertafore-side integration materializes in 2026-27.

How a challenger would attack it

Run the same product with clean hands. Comulate’s defensible ground was never parsing — LLM-era extraction has commoditized reading 13,000 statement formats, as Fintary and Eventual already show — it was AMS write-back depth plus the trust of the top-100 brokers. Both are precisely what the litigation destroyed: an injunction freezes new Epic-derived sales, a federal judge blamed the founders’ “own misdeeds,” and counsel has admitted customer defections and 20% layoffs. The challenger’s playbook writes itself: build the identical reconciliation engine on sanctioned access — a formal Applied partner agreement, or Vertafore-first — and sell into the pipeline Comulate can no longer close, leading every enterprise procurement conversation with compliance posture, audited security, and platform-blessed integrations, the exact axis where Comulate is radioactive. Ascend is already executing this via ReSource Pro’s distribution, converting the offshore-labor incumbent into the channel. The second attack is pricing structure: Comulate sells negotiated, high-ACV enterprise contracts justified by recovered commissions; a challenger prices on contingency — a percentage of commissions actually recovered — which transfers the proof burden and undercuts the quote-based model. Timing is everything: every month the Illinois docket grinds on, the exempt existing base (Lockton, Risk Strategies, IMA) is the only moat left, and moats made of customers under legal uncertainty erode on renewal dates.

Same playbook, new buyer

The core asset — ML that reconciles messy counterparty statements against a system of record and recovers leaked revenue — generalizes well beyond P&C brokerage commissions. The nearest shift stays in insurance but changes the buyer: carriers and MGAs face the mirror-image problem, reconciling producer payables and bordereaux across thousands of agency relationships, and the systems of record there (policy admin platforms, not Epic) have no Applied-style gatekeeper with an 80% share and a litigation appetite. Second, the wholesale and delegated-authority tier — wholesalers and program administrators sitting between retail brokers and carriers — runs double-sided reconciliation with even messier data and no incumbent software at all. Third, the same statement-to-ledger engine maps onto healthcare benefits commissions and even freight-broker settlement, both industries where PE roll-ups multiply incompatible books the way insurance consolidators do. The strategic lesson of Comulate’s collapse is itself the design principle: pick a vertical where the data choke point is fragmented or partner-friendly. Comulate cannot pursue any of these shifts — its capital, engineering, and founder attention are consumed by an existential Sherman Act case, and its brand now carries a court-documented conduct problem into every new-market sales cycle.

Sources and further reading

Capital history

DateRoundAmountValuationLead(s)
2022-2023 Seed (multiple tranches) ~$5M total (TechCrunch, Feb 2025); Crunchbase logs a $4M tranche in Dec 2023 Undisclosed Spark Capital, with Neo, Mischief (Zach Perret), Jonathan Crystal, Qasar Younis
Feb 2025 Series B (company skipped Series A) $20M Undisclosed BOND and Workday Ventures, co-leads; existing investors participated

Investors / owners: BOND, Workday Ventures, Spark Capital, Neo, Mischief (Zach Perret), Jonathan Crystal, Qasar Younis

Competitive set

  • Applied Systems — The Hellman & Friedman-backed AMS giant is simultaneously Comulate's most important platform dependency and its most dangerous rival. It owns Epic (which Comulate's antitrust complaint says holds 80%+ of enterprise brokers), the Ivans carrier-data network, and a native reconciliation product (Applied Recon) — and since November 2025 it has fought Comulate with trade-secrets litigation, revoked Epic access and, per Comulate's complaint, pressure on shared customers. Applied tried to acquire Comulate starting in 2023 and was refused.
  • Ascend — The best-funded direct rival: insurance payments and accounting automation with roughly $60M in equity plus large lending facilities (2021-23, company releases). Launched Direct Bill Automation in Sep 2024 covering the same statement-processing workflow, pairing it with agency-bill payments in one platform. Its Apr 2025 distribution deal with ReSource Pro gives it a services channel into agencies Comulate does not target — and makes it the natural landing spot for pipeline scared off by the litigation.
  • ReSource Pro — The status quo: a large insurance BPO to which brokerages have outsourced statement entry for two decades. Its Direct Bill Complete offering (with Ascend tech, Apr 2025) converts the offshore-labor incumbent into a tech-enabled competitor. Comulate's Heffernan case study — offshore reconciliation headcount cut to zero — shows exactly whose revenue it attacks.
  • Vertafore — The other AMS duopolist (Roper-owned; AMS360, Sagitta, BenefitPoint). Comulate integrates with AMS360 and BenefitPoint today, but Vertafore sells its own commission-tracking modules, and the Applied precedent shows an AMS owner can throttle a dependent vendor at will.
  • vBots / Synatic CommRecon — The RPA and data-integration tier: cheaper bot-based direct-bill reconciliation for agencies that will not pay enterprise prices. Houchens Insurance publicly upgraded from RPA to Comulate (company case study, 2024) — but these tools cap pricing below the top 100.
  • Fintary / Eventual — Younger AI-native commission-reconciliation startups chasing the same statement-parsing problem, mostly downmarket. Evidence that LLM-era document extraction is commoditizing the ingestion layer that was Comulate's original wedge.