Customer story
How Tryg unified decades of insurance data with Prisma
Tryg, one of the Nordic region's largest non-life insurers, handles over a million claims a year across data sources spread between countries and built up over decades. It uses Prisma to power Tryg 360, the in-house data-broker platform that pulls all of it into one model.
Prisma auto-generates database clients and GraphQL APIs straight from the schema, even for data models that run to 10,000 lines, so Tryg's developers iterate quickly and data becomes reachable for people without SQL expertise.
Customer
About Tryg
Tryg is one of the Nordic region's largest non-life insurance companies, covering private, commercial, and corporate markets and handling more than a million claims a year. To make its data accessible across the business, Tryg built Tryg 360, a proprietary data-broker platform, with Prisma as a core enabling technology.
Industry
Non-life insurance
Scale
Nordic's largest, 1M+ claims/year
Stack
CockroachDB, Kubernetes, Helm, Apache Kafka, GraphQL, Pal.js
Using
Prisma Client · Prisma Schema (generator API)
The challenge
Tryg's data had accumulated across different countries over decades, in systems that were never designed to work together. The result was a harmonization problem at scale:
Incompatible models. Data sources across countries used models built independently over many years.
Conflicting definitions. The same concept meant different things in different systems, creating redundancy and workarounds.
Slow, error-prone integration. Bringing it all together by hand was time-consuming and easy to get wrong.
Locked behind SQL. The goal was to make data accessible to everyone who needed it, including people without SQL expertise.
Why Tryg chose Prisma
Tryg chose Prisma as a critical enabler for Tryg 360 because it could generate the data-access layer from the schema and hold up under models far larger than most tools handle.
Clients generated from the schema
Prisma auto-generates database clients directly from the Prisma schema.
GraphQL APIs, auto-generated
Paired with Pal.js, the platform auto-generates GraphQL resolvers and type definitions for developers to work against.
Built for extreme complexity
Prisma handles very complex models and massive datasets, with schema files up to 10,000 lines and over a million characters.
“Prisma is a huge technical enabler for us.”
Artur Mrozowski
Data Engineer, Tryg
How they use Prisma
Tryg 360 is a data broker that streams live data from multiple sources into one place, without transforming it up front. Data synchronizes through a Time-Aware MirrorMaker built on Apache Kafka into a local Kafka cluster for selective loading, and lands in CockroachDB, which speaks the PostgreSQL wire protocol, as unified storage.
Prisma Client accesses CockroachDB, and from the Prisma schema the platform generates the database client and, with Pal.js, the GraphQL resolvers and type definitions. Whole environments spin up in one click through Kubernetes and Helm charts, so there's no manual configuration between a developer and a working environment.
generator client {
provider =
}
model User {
id String @id @default(cuid())
email String @unique
DateTime
posts Post[]
}
model Post {
String @id
author User @relation
}
Schema valid
The results
Tryg turned decades of scattered, incompatible data into one model developers can build on quickly.
Many sources, one schema
Multiple data sources unified into a single schema and data model.
One-click environments
Kubernetes and Helm provision whole environments without manual setup.
Faster development
Generating clients and APIs from code lets developers iterate quickly.
Our setup with Prisma enabled us to generate everything from code and ensure our developers can iterate very quickly.
Lasse Abelsen, DevOps Engineer at Tryg
Build on the same stack as Tryg
Tryg unified decades of insurance data on Prisma. See what it can do for your data model.
- Free to start, no credit card required
- Clients and GraphQL APIs generated from your schema
- Trusted by 500K+ developers