Products

Two products, one architecture principle: resolve the data before you build on top of it.

M1 Unify and M1 Assure are built on the same foundation — AI-native entity resolution and a privacy model that keeps every client's data inside its own boundary.

Master Data Management

M1 Unify

A single, governed view of customer, policy, and party data — resolved by machine learning instead of hand-written match rules that break every time a source system changes.

10+source systems typical
Weeksnot years, to first value
How it works
  • Ingests records from core, policy admin, billing, and third-party sources without a rigid canonical schema up front
  • Machine-learned matching identifies likely duplicates and relationships across systems, with confidence scoring on every link
  • Survivorship rules decide which source wins per field, with full lineage back to the originating record
  • A lightweight footprint — deployed alongside existing systems, not in place of them

Entity resolution

Probabilistic matching trained on insurance and financial services data patterns — names, addresses, and party relationships that rules-based MDM tools consistently miss.

Governance & lineage

Every merge, split, and survivorship decision is logged and reversible. Data stewards see exactly why a record looks the way it does.

Fits existing systems

Built to sit alongside Guidewire, Majesco, Oracle OIPA, and IBM IAA/IIW environments without requiring a core replatform.

Application areas
  • Underwriting triage — surfacing risk signals and missing information before a submission reaches an underwriter
  • Claims intake — structuring FNOL detail and flagging complexity or potential fraud indicators early
  • Rating support — explaining rating plan outcomes in plain language for underwriters and agents
  • Document intelligence — reading policy, endorsement, and loss-run documents into structured data
Insurance AI Applications

M1 Assure

AI applications built specifically for P&C and specialty insurance workflows — not a general-purpose model with an insurance prompt library on top.

1tenant per client, no exceptions
0client data used to train shared models
Privacy

Complete data isolation

Each deployment runs in a dedicated tenant or inside the client's own cloud environment. Outputs, prompts, and documents never train a shared model.

Domain fit

Insurance vocabulary, natively

Models and workflows are tuned on policy, claims, and rating concepts, so outputs read like they came from someone who understands the business.

Integration

Meets your core where it is

Pre-built connectors for Guidewire, Majesco, and Oracle OIPA-based cores shorten the path from pilot to production.

See M1 Unify or M1 Assure against your own data.

Most engagements start with a scoped pilot against a real subset of your book — not a generic demo environment.

Talk to us