- Qumis launched certified AI agents for commercial P&C coverage analysis, covering 16 lines of business including cyber, D&O, marine and workers’ compensation.
- The platform analyzes full insurance programs, including policies, endorsements and carrier forms, then produces citation-backed coverage conclusions tied to source policy language.
- Qumis is positioning the product beyond basic AI summaries, offering finished work products such as branded proposals, coverage checklists, Excel workbooks and hosted interactive dashboards.
Qumis, a coverage intelligence company for commercial property and casualty insurance, launched certified AI agents for coverage analysis across full insurance programs.
The agents act as line-of-business specialists, review policy documents, endorsements and carrier forms, then produce finished coverage analysis with citations tied to policy language. Qumis built the product for brokers, carriers, MGAs, TPAs, claims professionals and coverage counsel.
Users ask a coverage question or upload a complete insurance program. The file set might include thousands of pages across policies, endorsements and forms, according to Beinsure.
Qumis’ insurance-native orchestrator then assembles a bench of specialist agents across 16 coverage lines at launch, including cyber, D&O, marine and workers’ compensation.
Each specialist reviews its own coverage line at the same time. The system then combines the findings into one deliverable, with conclusions supported by citations to the source policy wording.
“AI in insurance has largely been about helping people do the same work faster,” said Dan Schuleman, Esq., co-founder and CEO of Qumis.
We built Qumis to make them more intelligent, so coverage professionals can see the whole program more clearly and make better decisions. Now every insurance professional can – with an entire bench of certified AI specialists at their disposal, on every file.
Dan Schuleman, Esq., co-founder and CEO of Qumis
Qumis said each specialist uses a documented, line-specific coverage framework. Qumis coverage counsel review and approve those frameworks before deployment. The company also validates the agents against representative coverage scenarios.
The system cites every conclusion back to the relevant policy language. When the documents do not provide enough evidence, the agents identify uncertainty rather than producing a forced answer. That matters in coverage work, where a polished wrong answer creates a bigger problem than a slow one.
Qumis also uses specialist agents to create final work products, not only coverage summaries. The system produces branded proposals, coverage checklists, Excel workbooks and live dashboards.
Users publish those dashboards as hosted microsites, replacing the static PDF proposals insurers and brokers have used for years.
Each engagement runs on Qumis Private Compute. The company describes it as a dedicated secure virtual computer, isolated from every other customer environment and powered by Qumis’ proprietary agentic infrastructure.
Give an agent a chat window and you get answers. Give it a computer and you get finished work. That’s the bet we made early with Qumis Private Compute, and it changes what we can promise professionals; not a summary of the file, but the work product itself.
Shiv Sinha, co-founder and CTO of Qumis
The launch arrives as claim severity rises and nuclear verdicts push coverage decisions into more expensive territory. E&O exposure has also raised the cost of an incorrect coverage determination for brokers, carriers and claims teams.
By reviewing entire insurance programs, Qumis gives coverage professionals a faster way to interpret policy terms, evaluate exposures and position coverage.
According to Qumis, the system brings the analytical discipline of coverage counsel together with market context, then turns it into work product at enterprise speed.
