Rate Repository
Akur8 is a cloud-based suite that helps insurers build and manage complex pricing structures, from modeling through deployment. Its rating solution, where actuaries convert models into actionable prices, was facing roadblocks in its expansion to the United States. The market operated differently, and so did its constraints. This led us to rethink our existing approach and design a new product tailored to American pricing needs.
Senior Product Designer
Design System Lead
Senior Product Manager
Technical Lead
Actuarial Data Scientist
4x Engineers
10 months
Role
Team
Duration
Key results
1M€ CaRR in first release year
01
Problem space
Navigating legislation
the United States regulates insurance pricing at the state level, meaning a nationwide insurer must update and report to over fifty different entities, multiple times a year.
3 core pain point defined the problem space:
Governance
Representation
the frequency of updates, combined with the regulatory system, results in insurers managing hundreds of pricing structures across dimensions of line of business, state, effective date, and customer types
to meet legal transparency requirements, the American industry favors table-based pricing structures, which can be complex to represent. The workflow-based tools available on the market handled this inefficiently (which included Akur8’s own Legacy Rating module).
Research
Research started with an advantage: regulatory insurance pricing filings are public domain in the United States. We were able to dive directly into real use cases across different lines of business, states, and company sizes. For the MVP, we focused on nationwide insurers with multi-coverage offerings such as home insurance, personal auto, and general liability.
Personas
We identified four key personas: decision makers (Underwriters & Product Managers), end users (Actuaries), IT, and support. For each, we mapped the core questions we needed to answer, their likely biases, their strengths as informants, and persona-specific hypotheses to validate. Different session types ran throughout development, sometimes in parallel: general workflow interviews, use-case-specific deep dives, and usability testing. We averaged two sessions per week across six months of product development.
The product centers on the Rate Order of Calculation, or ROC. Its core value is interactivity and centralization. Rather than manually assembling a rating plan from external tables and Python calculations, users bring their materials into the platform and reference them directly within the ROC. This creates a traceable, modular, and fully editable rating structure. The format itself deliberately mirrors the classic layout found in regulatory filings, so it feels immediately familiar to actuaries.
Every edit is tracked to provide full transparency on ownership, deployment status, and source data. Users can categorize inputs as factor tables or direct attributes and version each individual component of the ROC independently. A test section allows users to plug in a dataset and run the full calculation, ensuring IT receives a validated, bug-free rating plan.
The product generated $1M in committed annual recurring revenue before its official release, with $2M estimated by end of year.