Carrier cuts claims intake to 2 minutes

Insurance claims intake sounds boring, but it’s one of the most expensive little bottlenecks in the whole carrier stack. A new Mediar case example puts unusually concrete numbers on that problem: one regional property-and-casualty carrier cut intake and assignment from 30 minutes per claim to under 2 minutes, with claimed labor savings of $750,000 a year. That matters because a lot of “AI for insurance” pitches still live at the demo stage. This one is trying to sell a much simpler idea — stop making humans retype first-notice-of-loss data into old claims systems. ### What actually got faster? The workflow Mediar describes starts the moment a customer submits a claim. Instead of a claims rep pulling details from emails, web forms, or phone notes and then manually entering everything into the carrier’s system, the software routes the claim straight through intake and into adjuster assignment. Mediar says the handoff now happens in under 2 minutes instead of taking 30-plus minutes of staff time. (mediar.ai) ### Why is intake such a pain point? Because intake is where messy, unstructured information first hits the carrier. A burst-pipe claim, for example, arrives as some combination of text, attachments, timestamps, policy details, and urgency signals. None of that is useful until someone turns it into a structured claim record and sends it to the right adjuster. If that step is manual, every claim starts with delay, and every surge event — storms, floods, hail — turns the queue into a staffing problem. (mediar.ai) ### What is Mediar really automating? Turns out the pitch is less “magic AI adjuster” and more “finish the last mile.” Mediar’s own writeup makes that explicit: lots of document workflow tools can capture, classify, and extract data, but the real savings show up only when the parsed data is posted into the system of record. In insurance, that means the claim is not just read — it is actually created and assigned inside the carrier’s operational software. (mediar.ai) ### Why does the legacy-system angle matter? Because carriers do not run on clean greenfield software. Mediar says this workflow plugs into platforms like Guidewire ClaimCenter, Duck Creek Claims, customized Salesforce setups, and even legacy mainframe claims systems. That is the whole commercial hook. If a carrier already has a modern API-first stack, basic orchestration is easier. But many insurers still live in a world where the hard part is getting data into the screen that employees already use every day. (mediar.ai) ### How big is the claimed ROI? Mediar frames the example as a regional carrier handling 50,000 claims a year, with 93% time reduction and $750,000 in annual savings. On its face, that math is plausible for a high-volume clerical workflow — especially if the old process consumed half an hour of labor per claim. The catch is that this is vendor-published marketing, not an independently audited case study, and Mediar does not publicly name the carrier. (mediar.ai) So the useful takeaway is less “this exact number is proven” and more “buyers now have a specific benchmark for what intake automation is supposed to beat.” ### Why are people paying attention now? Because enterprise AI buying has shifted. A year ago, a lot of teams were still experimenting with copilots and summarizers. Now the pressure is on measurable workflow automation — fewer keystrokes, fewer handoffs, faster queue movement, cleaner data into the system of record. Mediar’s example lands neatly in that mood. It gives operators a before-and-after they can put in a spreadsheet. (mediar.ai) ### What’s the real lesson for carriers? Start where humans are acting like middleware. Claims intake is often exactly that — people reading one system and typing into another. That is a good automation target because the value is immediate, the workflow is repetitive, and the success condition is obvious: did the claim get created correctly and assigned fast? Mediar is basically betting that this narrow, boring, high-ROI wedge will sell better than broader promises about end-to-end autonomous claims. (mediar.ai) ### Bottom line The story here is not that AI suddenly “solved claims.” It’s that one vendor published a very crisp example of where automation can pay off right now — the first few minutes, before any real adjudication even begins. For enterprise buyers, that is often the part worth fixing first. (mediar.ai)