CLARUS.AI Request a demo
DeepRiskCommercial Lines

Underwrite complex risks with confidence. Make consistent decisions your book can defend.

DeepRisk transforms unstructured submissions and external data into validated proposals for more accurate risk assessment and pricing.

DeepRisk map view: a submission's objects on satellite imagery with policy PML, portfolio PML and peak accumulated PML, and the insurer's nearby portfolio within 200 metres
Exposure on the map: every object located, PML per location, and your existing portfolio within the radius. Demo environment, synthetic data.

Customer

BTA Baltic Insurance Company, member of Vienna Insurance Group
  • Lithuania, Latvia and Estonia

Why Clarus

Complex risks are still underwritten by hand.

The same risk, a different price, depending on who picks it up. Hundreds of objects per submission. Reinsurance treaty limits held in memory. Weeks to one proposal.

We built Clarus on decades of experience in insurance, across the Baltics and international markets. We saw how much complex underwriting still depends on manual work, even as technology transforms the rest of the industry. Clarus exists to change that. We put underwriters in charge: they define the rules and stay in control of decisions, while DeepRisk does the work behind them.

Start with a product. Grow with an experienced team.

DeepRisk · the product

Know the risk. Protect the book. Price it right.

Built for commercial lines: Commercial Property and Construction All Risks first, with further lines on the same foundation. Six modules, in the order a submission travels.

Know the risk

Every object located, enriched and measured before anyone quotes.

  • Submission IntakeAny RFP format becomes a structured, geocoded object list, with errors flagged at field level.
  • External InsightsMap, satellite and Street View, the neighbourhood, flood risk zones, the insured’s financial standing and the news.
  • Risk AccumulationPolicy PML, portfolio PML and peak accumulated PML, location by location, with your existing portfolio inside the radius.
DeepRisk object list: fifteen objects extracted from a submission with unique numbers, area, business type and structure, field-level error flags, the three PML figures and XLS download
Protect the book

Every proposal checked against the reinsurance treaty, product T&C and UW guidelines, with the rule and the value named.

  • ValidationA verdict on every proposal: insurable, insurable with conditions, or refer . Rules versioned and administered by your team.
  • Clause MatchingRequested clauses matched to your own standard catalogue, and flagged for wording only where nothing matches.
DeepRisk rule findings: three amber actions, each naming the wording or reinsurance programme clause behind it, with a suggested special-conditions clause, 11 checks run and nothing to ask the broker
Price it right

Every price argued from comparable policies, not a generic model.

  • Portfolio InsightsComparable policies from your own history: the median tariff, where this price sits, the segment loss ratio, and a tariff tester before you quote.
DeepRisk portfolio insights: 109 similar policies in the book, median tariff 0.025% against this proposal's 0.028%, loss ratio, average claim, and where this price sits in the typical range

What changes for the business

Faster to quote. Consistent to price. Safer to reinsure.

Time to quote

The submission is read, not re-keyed, so the broker gets an indication sooner.

Pricing consistency

The same risk priced the same way, whoever picks it up, with comparable policies on screen.

Treaty and wording discipline

Anything outside the reinsurance treaty, terms, guidelines or standard clauses caught before binding and named to its rule.

Risk accumulation control

Peak accumulated PML known for every location before the quote, with the existing book inside the radius.

The same team handles more customer requests. Your most experienced underwriters’ standard becomes every underwriter’s standard.

How it fits

Standalone first. Inside your core system when you choose.

One product, configured to each insurer: your rules, your clause catalogue, your languages, your core system. Your policy administration system stays the system of record, with nothing to migrate.

01 · Start here

Standalone. The full workspace, nothing to build.

Through single sign-on, with your rules and clause catalogue configured.

Works standalone

  • Submission Intake
  • External Insights
  • Validation
  • Clause Matching

Configuration, not an integration project.

02 · When you choose

Integrated. Two-way with your core system and warehouse.

When you want your own book in the picture, DeepRisk reads from and writes back to your core system and data warehouse. Risk Accumulation and Portfolio Insights work on your portfolio.

Embedded · builds on integration

The verdict inside your policy screen

Findings and actions where the underwriter already works, so nobody switches context.

DeepRisk widget inside a core policy system's coverages screen, showing an insurable-with-conditions verdict and required actions
The DeepRisk widget inside a core policy screen. Sample core system and synthetic data.

Wherever you want it. Standalone modules run on our infrastructure by default, or in your own cloud environment, run by us or by you.

  • Single sign-on, with access set by role.
  • Rules kept as versions. Every recommendation carries its evidence; every override is logged.
  • Built for European insurers: GDPR and DORA shaped the architecture.
  • Multi-language, from the broker’s submission to the screen.

Company

Insurance from the inside. Product at scale.

We built DeepRisk sitting next to underwriters, learning their work and where it could improve. DeepRisk is our product: built and owned by Clarus AI.

  • Clarus.ai, UAB
  • Vilnius, Lithuania
  • Founded 2023

Team

Dominykas Čeledinas

Chief Executive and Commercial Officer

20+ years in insurance software

Co-founded an insurtech acquired by TIA Technology (now Sapiens), an insurance core software provider, then grew TIA’s Lithuanian team to over 100 people, serving more than 70 insurers across EMEA.

LinkedIn

Linas Šimelionis

Chief Operating Officer

25+ years inside Baltic insurers

Board member and Director of IT and Operations at Lietuvos draudimas. Chief Administrative Officer at Gjensidige Baltic. Led a full core-system replacement for a Baltic insurer.

LinkedIn

Darius Matuliauskas

Chief Product and Technology Officer

25+ years building products

Co-founded a SaaS company acquired by Treatwell, then helped grow its B2B platform to tens of thousands of businesses across Europe.

LinkedIn

Request a demo

See a submission carried end to end.

We run the demo in our environment, on synthetic data, from the broker’s submission to a validated proposal.

We use these details to arrange your demo and follow up on it. See our privacy policy.