Documentation

Build on Owl.

Everything your team needs to deploy Owl on your claims systems: concepts, API reference, integrations, and the security model behind it.

01Documentation

Start where you are.

Eight categories, from the first connection to the model cards behind every output. Pick the one that matches the question in front of you.

01

Getting started

Provision your single-tenant VPC, connect your first claims system, and read your first file with Owl in under a day.

View
02

Core concepts

How Owl models the file, and why every output carries a span-grounded citation back to the page it came from.

View
03

API reference

Endpoints, request and response schemas, authentication, rate limits, and webhooks for every Owl product.

View
04

Integrations

First-party connectors for Guidewire, FINEOS, Origami, DXC, and Majesco, plus a generic ingestion path for the rest.

View
05

Security & compliance

The data-handling, tenancy, and audit model: SOC 2, single-tenant processing, and the controls behind every release.

View
06

Data taxonomy

The shared vocabulary Owl uses for claim entities, document types, and evidence, so your team and ours speak the same language.

View
07

Model cards

What Owl-1.5 and Owl-Bench are trained on, evaluated against, and validated for, with their published limitations.

View
08

Changelog

Every release, model update, and breaking change, dated and versioned, so you always know what shipped and when.

View
03Need a hand

Stuck on something the docs do not cover yet?

An Owl engineer can walk your team through the integration, review your architecture, or answer the question the reference left open.