Insurance AI
Q

Qumis

Qumis is an AI platform that commercial insurance brokers, claims teams and corporate risk managers use to read policies, check quotes and contracts, and prepare coverage positions and client deliverables. Coverage attorney Dan Schuleman and Shiv Sinha founded the Chicago company, and MTech Capital led its $4.3 million seed round in February 2026, with American Family Ventures joining.

Each policy is routed to one of sixteen specialists by line, from property and D&O to cyber and workers' compensation, which read grants, exclusions and endorsements together. A broker can compare a renewal quote with expiring coverage, check the attachment points in a layered program or test a live policy against a lease's insurance requirements. Every finding cites the page and clause, checked against the source document, and carries a reliance status from exploratory to reviewed. The same analysis feeds proposals, client microsites and stewardship reports, works alongside Outlook, Gmail, Google Drive and Slack, and reaches Claude and Copilot through an MCP connection in beta.

NFP, Brown & Brown, IMA Financial, CRC Group and Holmes Murphy are among the firms using it, and Qumis counts 150 organizations in production and 4.73 million document pages processed. In a survey of 109 users, 73% said it caught coverage issues they would have missed and 64% credited it with a client won, kept or grown.

Last VerifiedNovember 20, 2026
Compare Qumis with other vendors
Founded
—
Headquarters
Chicago, Illinois, United States
Website
qumis.com/
Categories
insurance-ai
Assessment

Capability Axes

Capability grades

15 of 15 axes rated · 8 graded A or B

AI Capability
AI Centrality
AA on AI CentralityThe artificial intelligence is the product. Remove the models and there is nothing left to sell.
Vendor Published

A policy or set of related documents goes first to a classifier, which routes it to the specialist for its line of insurance. The specialist reads grants, conditions, exclusions and endorsements and ties each provision to the coverage question, links documents across a layered placement, and spots material changes between a renewal quote and the expiring policy. The agents do the reading, comparison and first draft that make up the work brokers, claims teams and risk managers buy the product for, so without them there is nothing left to sell.

Autonomy and Oversight Model
AA on Autonomy and Oversight ModelWhat the system runs alone, what constrains it, and how a person checks it are all published: modes, thresholds, sampling or audit controls, and the route a case takes to human review.
Vendor Published

Every output carries a reliance status, so a reviewer can tell an exploratory finding from a reviewed one, and each finding points to a citation checked against the original page and clause. When policy language is ambiguous, the product flags it and routes the close call to the professional responsible for the file, and in a layered program a finding names the attachment point or provision behind the issue. The coverage team decides what to do with every finding, with the source language in front of it.

Model Risk Management and Transparency
BB on Model Risk Management and TransparencyReal transparency mechanisms are published, such as per alert explainability, confidence scoring or split testing, without the validation package or supervisory mapping behind them.
Vendor Published

Coverage findings cite the clause, endorsement or other source that controls the result, and the citation is checked against the actual document rather than generated as text. A coverage ontology connects exclusions, endorsements, sublimits and claims to the wider program, so a user can inspect the links when a quote departs from expiring terms or two layers do not line up. Reliance status separates exploratory findings from reviewed ones. A validator can reconstruct any single answer, but there is no supervisory model risk mapping or measured validation of the specialists.

Operational and Outcome Evidence
BB on Operational and Outcome EvidenceVendor aggregate claims with real figures, or audited scale disclosures from a publicly listed company.
Vendor Published

NFP adopted Qumis for risk and policy analysis, Connor Love at IMA Financial describes using it to find discrepancies in layered programs, and Kapnick Insurance and Cruser Mitchell have named references too. Qumis counts 150 organizations in production, 4.73 million commercial property and casualty pages processed, and seat growth of four to six times at its largest brokerage deployments over the past year. In a survey of 109 users, 73% said it caught a coverage issue they would have missed. The one timed result, a comparison cut from four hours to five minutes, comes from an unnamed account manager.

AI Safety and Data Stewardship
BB on AI Safety and Data StewardshipA categorical stewardship commitment is published without the retention schedule or the engineering detail behind it.
Vendor Published

Private Compute keeps customer documents on dedicated infrastructure and never uses them to train models, and the service as a whole carries a no training on client data commitment. Customers manage access through single sign on and admin controls, and the trust center lists deletion of customer data when a client leaves. Reviewed decisions do improve the coverage intelligence, and how that feedback is kept apart from other clients' work is not explained. No model provider appears among the subprocessors, so retention terms at the model layer are not stated either.

Regulatory and Compliance
GLBA and Data Privacy Posture
BB on GLBA and Data Privacy PostureA substantive privacy document that reaches the product itself, short of the subprocessor list or the full data handling detail.
Vendor Published

The privacy policy covers information collected through the website and applications, including account, contact and transaction data, gives a route to close an account, and allows records to be kept afterward for service and legal purposes or in archives. Its California supplement excludes processing in the role of a service provider. The trust center lists GDPR and HIPAA attestations and a control for deleting customer data when a client leaves. A retention schedule for a broker's policy and claims files is not given.

Security Certifications and Trust Center
BB on Security Certifications and Trust CenterA recognized certification named in the vendor’s own material without the artifact, or with a scope or renewal question the buyer has to raise.
Vendor Published

Qumis completed its SOC 2 Type II examination in September 2026, after a Type I in July 2024, and its trust center lists the 2026 report beside HIPAA and GDPR attestations and an EU AI Act memo. The report is shared under NDA rather than on open download. Product controls include dedicated infrastructure for documents, single sign on and workspace administration, and the trust center tracks encryption, penetration testing and disaster recovery controls.

Regulatory Status and Licensure
CC on Regulatory Status and LicensureThe regulatory position is unstated. Most vendors in this index are technology suppliers and being unlicensed is the correct posture, so this grade records silence about the posture, not a missing license.
Vendor Published

Commercial insurance teams use the product to inspect coverage and draft advisory work, comparing policy language, claims evidence and program structure so a broker or coverage specialist can review a position, and to show which contractual insurance requirements are unmet or ambiguous. It is analysis software and holds no producer license, regulatory registration or filing role of its own. The trust center carries an EU AI Act compliance memo, which sets out the company's own reading of the act rather than any regulator's admission.

AI Governance and Bias Disclosure
CC on AI Governance and Bias DisclosureResponsible artificial intelligence committed to in policy language with no evaluation behind it, on a product whose bias surface is modest.
Vendor Published

Practicing coverage attorneys help build the coverage specialists. The product flags ambiguous language for professional review and records the reliance status of each output, which settles who judges a coverage reading and how ready it is for advisory work, and keeps a preliminary extraction apart from a reviewed conclusion during a claim or renewal. An EU AI Act compliance posture statement sits in the trust center. There is no fairness evaluation, which matters less here because the specialists read policies rather than score people.

AI Liability and Recourse
CC on AI Liability and RecourseMechanisms that enable challenge, such as audit trails and source traceability, with nothing standing behind the output and no route for the person affected.
Vendor Published

A broker can check a coverage finding by opening the cited page and clause, and the system flags ambiguity and routes the finding to a professional with its reliance status visible, so a firm can catch a wrong reading before acting on it. The website terms disclaim warranties, including on site accuracy, and cap liability tied to the site at $50 where the law allows. No guarantee covers a loss from an incorrect coverage position.

Integration and Deployment
Model Supply Chain Disclosure
CC on Model Supply Chain DisclosureThe architecture is described and no provider is named.
Vendor Published

A classification and routing step sends each policy to a coverage specialist, which works with a coverage ontology linking the terms of the program and cites the original file. The public subprocessor list names Amazon Web Services as the cloud provider, with Google Workspace, Vanta, Sentry, GitHub and Rippling for internal tools, and lists no model provider. Whether the specialists run on models Qumis built, on a hosted foundation model or on both is not stated.

Core Systems and Integration Depth
CC on Core Systems and Integration DepthIntegration claimed through standards or connectors with no system named and nothing to verify.
Vendor Published

A broker loads policy and claims material into the workspace, keeps it in the Vault and exports or shares the finished report, and the same findings can feed a branded proposal or client microsite. The product works alongside Outlook, Gmail, Google Drive and Slack, and its MCP connection, in beta, brings the coverage analysis into Claude and Copilot. Integrations with agency management systems are in early access with design partners, so no broker management, policy administration or claims system is connected in production today.

Deployment Model and Data Residency
CC on Deployment Model and Data ResidencyCloud only with nothing stated, which is the category norm.
Vendor Published

Teams use a hosted application, running on Amazon Web Services, with document upload, the Vault and workspaces for reports and client material. Private Compute puts a customer's documents on dedicated infrastructure with customer access controls. The hosting region, and any choice of region, are not stated, and there is no on premises option.

Commercial
Commercial Transparency
CC on Commercial TransparencyNo price is published and engagement runs through a demo form, which is the norm in this index.
Vendor Published

Qumis sells through a demo that runs one job on the prospect's own documents in thirty minutes. Firms can buy it for policy checks, claims work, proposals, contract review, program design and stewardship reports, and carriers, managing general agents and administrators have a separate route to work with the company directly. Price per seat or document, an enterprise minimum and implementation costs are quoted rather than listed. Reliance status labels the review of an output and is not a subscription tier.

Institution and Segment Coverage
AA on Institution and Segment CoverageThe financial segments served are named and each carries its own maintained material, whether the coverage is broad or deliberately narrow.
Vendor Published

Qumis is built for commercial property and casualty. National brokerages, regional firms and wholesalers such as CRC Group compare program structures and coverage terms, claims specialists assemble coverage positions with the policy language attached, and client risk teams check how the current program meets a contract's requirements. Coverage attorneys are a separate audience, with insurance law firm Cruser Mitchell among the references.

Use cases run from policy checking and contract requirements to proposals, program design and stewardship reports, and carriers and claims organizations can arrange their own workflow with the company.

Alternatives to Qumis

Other products that do the same kind of work as Qumis, and what each one does.

Powerbroker AI is document analysis software for commercial property and casualty agencies, based in San Francisco and founded by Marcus Gallagher, its CEO, and Patrick Cooney, its president.

Ennabl sells ennabl.AI, an AI workspace for insurance agencies and brokers, managing general agents, wholesalers, reinsurers and agency networks, from Greenwich, Connecticut.

Outmarket AI automates the document work that fills a commercial insurance brokerage, connecting into the agency management system of record and reading carrier quotes, policies, applications and loss runs to produce comparisons, coverage gap analyses, policy checks confirming that what…

InsurGrid gives property and casualty agencies a way to collect, verify and compare customers’ existing insurance coverage.

Bold Penguin runs a digital exchange for commercial insurance where retail agents and brokers, carriers, MGAs and carrier owned distributors place business.

Sure Verify is Sure’s insurance verification software for rental property managers and the software providers that serve them.

Commercial

Pricing

Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.

Entry Price Pricing Basis Data Protection Terms Implementation Source
Quoted after a demo on the prospect's own documents; no list price.
Qumis sells a hosted coverage intelligence service through demos. It supports policy review, claims coverage, contracts, program structure, proposals and stewardship in the same workspace. Qumis Private Compute uses dedicated infrastructure for customer documents, with single sign on and administrative controls, and Qumis excludes client data from model training. A HIPAA attestation and a GDPR attestation sit in the trust center. Implementation charges are not listed. A product demonstration runs one use case on the customer’s documents. Carriers and claims organizations can work with Qumis directly on deployment. Vendor Published

Rates per seat, per document and by account are not listed. Reliance status describes review of an output, from exploratory to reviewed. It is part of the product workflow.

Contact us

Found a vendor we missed? Have feedback on the index? We’d love to hear from you.

AI FinTech Index

The AI FinTech Index is an independent index that tracks changes to AI vendors in financial services. It holds 797 vendors across banking, lending, insurance, wealth, capital markets and financial crime compliance, each graded on the same 15 capability axes from public sources. No vendor pays for inclusion, placement, or rating.

Index Status
Last index update
November 21, 2026
The AI FinTech Index is an editorial reference, not a regulatory body. Vendor data is verified against published sources and public regulatory filings. Figures labeled “Estimated” have not been confirmed by the vendor. See the Methodology page for evaluation standards and limitations.
© 2026 AI FinTech Index
3801 N Capital of Texas Hwy, Ste E240 · Austin, TX 78746