- October 15, 2026
- v8.25.2
October 15, 2026
New Requirements PDF
Risk Requirements PDF is available for initial renewal outreach in the case intelligence menu.
Changelog
Risk Requirements PDF is available for initial renewal outreach in the case intelligence menu.
Case chat has been folded into Halley, so there is a single conversation on a case rather than two. When Halley is close to a match but not certain, it now asks rather than deciding quietly.
Filter options can be selected and cleared in bulk, tasks can be filtered by name, and the policy compliance report adds a Policy Start Date column with filtering. Report headers have been reworked for consistency across reports.
Halley now presents the policy structure it intends to build before writing any of it, so you can see how it has read a program and change it rather than correct it afterwards. Its status in a case is clearer while a run is in progress.
Saved reports open in the state you left them, and the tasks report loads faster on large books.
Any report you have filtered, grouped, or configured can now be saved and returned to, so a quarterly view or a weekly book review is one click away instead of rebuilt each time. Saved reports live in the left menu under Saved reports, showing type, owner, visibility, and when each was last updated. Keep a report private or share it with everyone in your organization, and view, duplicate, transfer, or delete it from the same menu.
Cases and policies can now be placed on hold to mark work as paused. Putting a case on hold cascades to its active policies, and coming out of hold either releases the record back to the status it held before.
New endpoints place cases and policies on hold and release them, so a loan can be paused and fully resumed through the API alone.
Comps and pricing data now highlight medians rather than averages, giving a truer read on where the market is actually clearing. A three-month window has been added to the date filter, and carrier results now align with comp group size.
The tasks report now supports export, and can be grouped by Insured Group so escalation performance can be reviewed by sponsor. Task reports become generally available September 11th.
Requirement math functions can now use effective date, expiration date, current date, and expected closing date, so requirements can be written against timing rather than values alone. Release includes additional WIM fixes and optimizations.
We introduced a task report covering escalation work across your book. It separates escalations handled in-window from those handed off, and shows how many tasks are open, actioned, overdue, and completed, alongside how long a task takes to action on average. The report is available by request and becomes generally available on September 11th.
Building and maintaining requirements takes fewer steps with bulk tools and built in dry runs. Risk profiles are searchable, and a new matrix view allows for quick audits of the most important coverages.
Re-uploading a document now records it as a new version of the original rather than adding a second copy alongside it, so a document's history reads as one thread. Release includes additional minor improvements to case agent.
Additional filtering now available on pricing and compliance reports.
You can now choose which widgets appear on your home screen and arrange them to suit how you work. A portfolio manager watching compliance across a book and a broker tracking renewals have fully independent screens. Choose from our library of widgets to get started.
The case agent accepts mixed document uploads, so a bundle containing several kinds of document can go in as it came to you. It can also add policies to a case that has already been approved post-close, and it now surfaces every change it makes, so its work is reviewable rather than assumed.
A new report lays limits and deductibles out as a coverage-by-peril grid, so you can see at a glance where your portfolio is at risk and what specific coverages and perils are affecting your books compliance.
When the agent stops on something it cannot resolve, you have a way forward. Failed runs are restartable, a run waiting on input stays parked when a retry comes in, and the reason a run stopped is visible in the product instead of requiring a conversation with our team.
Creating a case manually is a first-class choice again, alongside agentic upload, for the times you would rather enter the details yourself.
Pricing now reads as rate on line per $100.
New accounts can now start a 14-day free trial of Premium and continue on the Free plan when it ends, with no commitment up front.
We shipped WIM 3.0, a smarter engine with an improved ability to parse complex insurance structures like layered towers, blanket programs, and shared limits spanning multiple policies. Read more about our models on our Labs page. WIM 3.0 is now widely available.