Services run as lightweight Linux containers that scale up for busy hours and down overnight, instead of fixed machines sized for the busiest day of the year.
Linux containers, with no Windows server licensing in the compute bill
Stateless services that scale by adding instances
Our own full environment runs in Azure containers for around $200–$300 a month
Development and test environments sized small, and paused when idle
Designed to scale with demand, not with the calendar
Open source
Standard parts, not a proprietary stack
The platform is built from open-source building blocks, a standard scripting language for rules and Linux containers. There is no proprietary rule language and no separate authoring product to license.
A standard scripting language for rules and extensions
Open-source runtime and cache components
Tooling your teams and your auditors already know
Web-based
Nothing to install, nothing to host per person
Product teams, underwriters and customers all work in the browser. There is no desktop authoring software and no virtual machine per user, so there is no fleet of workstations to buy, patch and pay for.
No virtual machines for development either: our own team builds on ordinary laptops and in the browser workspace
One browser workspace for product, actuarial, QA and compliance teams
Updates ship once, for everyone
Customer and agent portals on the same engine
MACH architecture
Pay for the parts you use
Microservices, an API-first design, cloud-native delivery and a headless front end mean each part scales on its own and each channel plugs in through the API.
Scale the pipeline and the runtime independently
Add a broker portal or embedded channel without reworking the platform
Change one service without retesting everything
No lock-in
Your cloud, your AI, your choice
Standard containers and open-source components are designed to run on any cloud. The platform runs on Azure today and can run on AWS too, on Windows, Linux or macOS. It works with the database you already use: SQL Server, PostgreSQL, MySQL or Oracle. AI providers are a configuration choice, and documents can live in cloud storage or in the database.
Switch between Azure OpenAI, Azure AI Foundry and Anthropic Claude
Runs on Azure or AWS, with the same containers
Runs on Windows, Linux or macOS, in the rules engine as well as in the interface
Your database: SQL Server, PostgreSQL, MySQL or Oracle
Pluggable document storage
Open, relational product data you can export
People
Build products with analysts, not specialists
Product work happens in the browser workspace. Business analysts, QA and prompt engineers build and maintain products, with AI agents drafting and people approving. A standard scripting language means less to learn and a normal hiring pool.
No proprietary language to train people on
Analysts approve, AI agents draft, QA tests every rule
Shorter onboarding in one workspace
What it adds up to
A smaller bill, a leaner team
A smaller infrastructure bill, no per-seat authoring fleet and a leaner team to run it. We would rather show you a cost model built on your volumes than quote a generic number.
