Openkoda: Most Customizable Insurance PAS

The center of gravity in insurance has shifted toward shipping products fast.

MGAs and insurtechs lead here by design. They price, underwrite and bind in niches that larger carriers can't move on quickly, and that edge only holds if their software keeps pace. Carriers feel the same pressure as specialty lines, embedded distribution and parametric cover grow.

The problem is that most policy administration systems were built for standard products. When a product doesn't fit the vendor's data model, or a pricing change has to wait for a release, the system decides what the business can sell.

Openkoda is built for the opposite situation: a complete policy administration system that the insurer can reshape, quickly and safely, as the business changes.

What Openkoda is

Openkoda is a highly customizable policy administration system (PAS) for insurers, MGAs and insurtechs. It runs the whole insurance lifecycle on one platform and one data model: product and rating, underwriting, policy administration, claims, billing, documents, bordereaux and reporting.

A bind, an endorsement, a first notice of loss and a payment all update the same record. Underwriting, servicing, claims and finance teams work from one source of truth instead of reconciling separate systems.

It runs fully managed on Openkoda's cloud or on-premise in your own environment, and your configuration and data can be exported at any time.

Customizable at every level

Openkoda's defining trait is how much of it your own team can change, and how safely.

• Products and pricing. Coverages, risk attributes, limits and deductibles are configuration. The rating engine uses multi-dimension factor tables and formulas, tested on sample risks and rolled out by effective date.
• Underwriting rules. Auto-approve, refer and decline rules are authored by the underwriting team, and changes apply to the next submission without a release.
• Workflows. Quote, referral, issuance, renewal and claims processes are drawn on a visual workflow canvas and simulated before they go live. Even the buttons on a policy screen are configurable: a new action, such as emailing documents to a client, is a short workflow that appears for the whole team once activated.
• Forms, documents and emails. Quote forms are built in a visual form builder with conditional questions. Policy documents are Word templates with placeholders, and emails are designed in a visual editor.
• Screens and dashboards. Dashboards are built per role from a set of tiles, or described in a sentence and generated.
• Roles and access. Fine-grained, role-based access down to field level, with several brands or organizations on one installation and their data kept separate.

Two things make this level of customization safe to use in production. First, every product change creates a new version, so policies already in force keep the terms they were bound under while new business moves to the current version. Second, every change is logged in a full audit trail.

And where configuration isn't enough, such as a proprietary pricing model or an unusual partner integration, the platform can be extended on the same foundations, by your developers or by Openkoda's team.

AI that drafts, people approve

AI runs across the platform, always as a draft for a person to review:

• AI Product Builder: describe a product or a change in plain English, or upload a requirements document, and review a before-and-after change-set covering coverages, rating, rules, forms and workflow before anything is applied.
• Reporting AI: ask questions about your book in plain English and get a report, with the generated query shown. Only the database schema goes to the model; the data stays in your system.
• Underwriting copilot: highlights the risk and pricing factors that matter on a referred case. The underwriter decides.
• AI document reading: extracts details from submission documents and attachments for a person to confirm.
• MCP server: lets assistants such as Claude work with the live book, limited to the user's own permissions, with every action confirmed and audited.

Key features

• Policy administration: quote, bind, issue, endorse, renew, cancel and reinstate, with a 360-degree view of each policy, customer, claim and billing account
• Underwriting: straight-through rules, a shared referral queue and a configurable underwriter workbench
• Claims management: guided FNOL from any channel, configurable routing, reserves, payment approvals, recoveries and fraud flags
• Billing and payments: payment plans, automated invoicing and dunning, card payments and commission tracking
• Distribution: branded quote-and-buy forms, embeddable widgets for partner checkouts, and client and broker portals
• Bordereaux: premium and claims bordereaux generated on each program's schedule, in the Excel and CSV formats carriers expect
• Team collaboration: chat, tasks and notifications attached to policies, claims and submissions
• Integrations: REST APIs, webhooks, a submission API on every form and the MCP server, plus an ACORD-aligned data model

Who it's for

Openkoda isn't trying to be the right answer for everyone. It's built for insurance businesses that need a complete core but write products, or work in ways, that don't fit a standard template.

By operating model:

• MGAs and coverholders running programs for several carriers, with binding authority checked at quote time, commissions tracked and bordereaux generated automatically
• Carriers and P&C insurers running personal, commercial and specialty lines on one core, or replacing a legacy system one line of business at a time
• Insurtechs and embedded insurance providers selling inside partner journeys, priced in real time
• Brokers and agencies that need a client view, renewals, commissions and self-service portals
• TPAs and claims administrators handling claims per client program, with payment controls and SLA tracking
• Mutuals, captives and reinsurers that want modern software without handing control to a vendor, including takaful operators with their own fund structures

By line of business, one configurable model already runs motor, property, cyber, travel, marine, pet, aviation and life, including multi-line and package products. It suits specialty and non-standard products in particular, because the data model, rating and rules are yours to define.

Two published projects show what that looks like in practice. Core Travel Insurance went from implementation to production in eight weeks with its group travel platform, including migrating existing clients. SkyGuard runs long-term life cover and specialty aviation P&C on one Openkoda instance, connected to its own AI rating engine, after about three months of delivery.

Freedom and predictable costs

Customization only matters if you keep control of what you build. Openkoda runs on its managed cloud or on-premise in your own region, and your configuration and data can be exported at any time.

Pricing follows the same logic. Plans are published and flat, with unlimited users and no per-seat, per-policy or percentage-of-premium fees, so growing your team, adding products or entering new markets doesn't trigger a renegotiation.

For teams that need to launch and change products quickly without boxing themselves in, that's the combination that matters: a complete PAS, the freedom to reshape any part of it, and costs that stay predictable as the business grows.

Here's the AI Product Builder configuring a new product from a requirements document: