Terrene Labs joins the FRISS family! Find out how together we will take Commercial Underwriting to the next level! Read more

NAICS Classification by Underwriting AI

Could half your books be coded wrong? Industry misclassification for SIC/NAICS is at 52 percent*. Are you accepting premium leakage, undesirable risks, incorrect bureau reporting or damaging your reinsurance relationship?
With a 52% error rate, misclassification greatly impedes carrier performance.

 We understand the criticality of gathering information that informs the insurance carrier of the true nature of operations at the businesses they want to insure. We also report on businesses doing multiple operations that traverse coverage boundaries allowing for the underwriters to address premium leakage.    

 We let you focus on the core: providing coverage for identified operations, and we do the chore: gather insight on specific operations that the businesses are involved in so that the coverage and premiums are aligned. 

“*5-year study by Verisk”

Underwriting AI leads to the insurers’ book reflecting the actual risk.
Average appetite of most carriers with 1057 NAICS codes


The number of classes an experienced underwriter can be deeply knowledgeable on


Estimated premium leakage in the first year for BOP policies


Leverage the combined knowledge of the multitudes for broader underwriting insight.
Machine Learning

Automate processes with our intelligent and ever-evolving classification engine.

  • Entity verification
  • Licensing/Accreditation validation
  • Quick classifications
Better Data

All available data is sourced, blended and verified to provide confident insights.

  • Dynamic sourcing
  • Proprietary blending
  • Automated insight matching

Risk classification can and should be clear cut. Code each risk with confidence.


Data Harvesting

Data is gathered from limitless sources, expertly blending and extracting crucial factors into actionable insights.

Dynamic Determinations

Going beyond data aggregation, key information is derived from structured and unstructured content to enable better and enriched decision making for evolving risks.

Image Analytics

Data is mined to produce actionable, accurate context that previously went unnoticed.

Natural Language Processing

Unstructured text found on the web and social media is analyzed, alerting you to conditions that could affect your decisions.

Machine Learning

Risk classifications are inferred with speed and accuracy by our proprietary risk engine by leveraging multiple, continually trained models.

Frictionless Experience

Dynamic risk insights are delivered in real time, providing a joyful experience for your customers, agents, brokers, and underwriters.

Trusted by Visionaries and Market Leaders