Markwell
Schedule a demo

Risk Intelligence API

Real-time risk intelligence for safer underwriting and claims decisions

Built for the LATAM insurance market. A REST API your team can integrate in a sprint.

Schedule a demoRead the docs
Portfolio · Workers' CompLive

Active policies

10,607

Loss ratio (12m)

67.4%2.1pp

Avg. score

718

Risk distribution

Low60.4%
Med27.2%
High12.4%
PolicyScorerisk
  • POL-44218847Low
  • POL-44217612Med
  • POL-44216318High
  • POL-44215782Low

The problem

Underwriting and claims are still decided in the dark

Three friction points that erode margin on every policy cycle.

Manual rules that don't scale

Underwriting still relies on legacy criteria and underwriter intuition. The result: inconsistent decisions and elevated loss ratios.

Fraud detected after payout

Anomalous patterns surface only after the money is out the door. Reactive detection costs margin and reserves.

Portfolio visibility lagging by weeks

Exposure and loss-ratio reports arrive late. By then, risk concentration is already a fact, not a forecast.

How it works

One API. Three modules. Integrate in days.

REST over HTTPS. No proprietary SDK. No extra infrastructure.

Underwriting Scoring

Real-time risk score 0–1000 per policy, with explainable factors.

POST /v1/score

Claims Antifraud

Anomaly detection over every incoming claim.

POST /v1/claims/screen

Portfolio Analytics

Loss ratio, exposure by line of business, and geographic distribution, up to date.

GET /v1/portfolio
RequestPOST
POST https://api.markwell.ai/v1/score
Authorization: Bearer mw_live_•••
Content-Type: application/json

{
  "policy_type": "auto",
  "applicant": {
    "age": 34,
    "region": "CR-SJ",
    "history_years": 7
  },
  "coverage": {
    "amount": 8500000,
    "deductible": 250000
  }
}
Response
{
  "score": 782,
  "tier": "low_risk",
  "factors": [
    { "name": "clean_history",   "weight": 0.34 },
    { "name": "region_low_loss", "weight": 0.21 },
    { "name": "age_band_25_40",  "weight": 0.18 }
  ],
  "latency_ms": 412
}

Example uses synthetic data representative of the Costa Rican market.

Use case

Beyond rules: the model finds the risk that slips through

Synthetic data — Demo

8 pp

Projected loss-ratio improvement

Applied to a synthetic portfolio of 10,607 active policies in Workers' Compensation, Markwell's scoring model identifies high-risk policies that traditional rules miss.

10,607

Policies analyzed

12 months

Projection horizon

Synthetic data representative of the Costa Rican market. Projections are illustrative, not guaranteed.

Who it's for

Built for every actor in the insurance chain

Same API, different value depending on where you sit in the policy lifecycle.

Insurers

Increase consistency and lower loss ratio across high-volume underwriting.

Insurtechs

Ship digital products with scoring from day one, without standing up a data team.

Brokers and MGAs

Triage quotes before committing capacity or going out to market.

Reinsurers

Analyze concentration and exposure across the portfolios you take on cession.

Security & compliance

Built on infrastructure your auditor recognizes

Encryption in transit, data minimization, and AWS-native operations. No surprises in your security review.

AWS us-east-1

Serverless infrastructure on a production AWS region with high availability.

TLS 1.2+ with ACM

Certificates managed by AWS Certificate Manager. No traffic in the clear.

API key — OAuth 2.0 on the roadmap

Key-based authentication today. OAuth 2.0 flows for enterprise integration in 2026.

Data minimization

We ingest only the fields needed for scoring. No unnecessary PII.

Costa Rica Law 8968

Data handling aligned with Costa Rica's personal data protection law.

SOC 2 on the roadmap

SOC 2 Type I targeted for 2027. Progress reports available under NDA.

Pricing

Pricing built for every stage of your business

We size volume and SLA together. No surprise numbers.

Starter

Early-stage insurtechs and brokers

  • Per-call pricing
  • Free sandbox
  • Low monthly minimum
  • Self-serve onboarding

Growth

Mid-size insurers

  • Monthly subscription with included volume
  • Per-call overage
  • Standard SLA
  • Dedicated support channel

Enterprise

Tier-1 insurers and reinsurers

  • Annual contract with custom volume
  • Guaranteed SLA
  • VPC or on-prem deployment
  • Assisted onboarding

Frequently asked

What teams ask before they sign

How long does integration take?

First working scoring call takes 3 to 5 days. A production integration with your policy system typically lands in one standard sprint.

What data do you need from us?

For baseline scoring, just the applicant and coverage minimums. To retrain on your portfolio, historical policy and claims data.

Is my customer data used to train shared models?

No. Each customer has their own model. Your data does not train other customers' models or our public models.

What regions and languages do you support?

Costa Rica as the primary market in 2026, expanding to Panama, Guatemala, Colombia and Mexico through 2026-2027. API and docs in Spanish and English.

How is pricing structured?

Three tiers — Starter, Growth, Enterprise. Per-call in Starter, subscription with included volume in Growth, annual contract in Enterprise.

Can we run Markwell in our VPC or on-premise?

Yes, on the Enterprise tier. Deploy in your AWS VPC or on-prem for flows with data-residency requirements.

What's your SLA?

99.9% uptime is the standard target on Growth. Contractual SLA with financial commitment on Enterprise.

How do I get access to the sandbox?

For now, by scheduling a call. We activate sandbox credentials for each design partner during onboarding.

Let's talk

Let's talk about your portfolio

I'll get back to you within 24 hours. Founder-led, no middlemen.

Schedule a demo
Mario Perez

Mario Perez

Founder, Markwell

San Jose, Costa Rica