Sytrex
AI-Powered Co-Pilots for Smarter Insurance and Reinsurance Operations
At a glance
- Vision / Challenge
- Design and build Steve, an AI-powered Co-Pilot for insurance and reinsurance brokers, from the ground up, turning a fragmented, manual process of scanning policies, extracting data, ranking reinsurers, and chasing communications into one intuitive, intelligent platform.
- Impact
- Policy scanning, data extraction, reinsurer ranking, and communication, previously four separate manual tasks spread across tools and inboxes, unified into a single guided workflow inside Steve, letting brokers move from policy to placement with far fewer manual touchpoints.
- Key Services
- Product Strategy & Design · Artificial Intelligence
- Industry
- Insurance / Reinsurance
- Technologies/Platforms
- Web-based AI Co-Pilot application · Hi-fi prototyping · Design system / component library
Steve: an AI Co-Pilot that unifies policy scanning, data extraction, and reinsurer ranking into one guided workflow for insurance and reinsurance brokers.
Brokers doing everything except brokering
Sytrex came to Triuum with a clear commercial goal: streamline broker operations and boost deal flow. But underneath that goal sat a genuinely hard workflow problem. Brokers were spending the bulk of their time on work that wasn't actually brokering, manually scanning policy documents, extracting key data points, cross-referencing and ranking reinsurers, and following up on communications by hand.
None of that friction was going to be solved by a better-looking dashboard. The real brief was to rethink how a broker moves from a raw policy to a ranked, ready-to-send placement, and only then design the interface that makes that path effortless. That meant the product needed to do real cognitive work on the broker's behalf, not just display information faster.
From policy to placement
The team mapped the broker's end-to-end workflow, from receiving a policy to closing a deal, identifying exactly where manual effort was highest: reading dense policy documents, pulling out the data that mattered, and manually comparing reinsurers against that data.
These findings shaped Steve's core architecture: an AI Co-Pilot structure built around three connected capabilities, with automated communication layered on top rather than bolted on as an afterthought.
The chosen direction was developed into a full UX/UI for the web application, giving Steve a consistent, trustworthy interface for handling something as sensitive as insurance and reinsurance data.
A co-pilot, not a black box
The clearest strategic call was treating Steve as a co-pilot, not an automation black box. Brokers weren't looking to hand off decisions entirely, they needed to work faster without losing the judgment that makes them good at their job. So instead of designing four disconnected tools (a scanner, an extractor, a ranking engine, a messaging tool), Triuum designed one continuous flow: scan a policy, see the extracted data surfaced clearly, review an AI-ranked list of reinsurers, and act on it with communication triggered from the same place the decision was made.
That single design decision, made early and carried through every phase, is what turned four separate manual tasks into one intuitive platform.
You may also like
Let's build your next triumph
Tell us about the challenge you're facing. We'll bring the team, the thinking, and the craft.