Soteris
Soteris is a company based in United States. Builders in Fintech has recorded 1 equity round for Soteris: a seed round of $8M on 22 September 2026.
Added ·
- $8M
- Seed · $8M · 22 September 2026 · led by Spider Capital
Timeline
Seed · $8M · Spider Capital (lead), Amplify Partners, DCVC, Intact Ventures, Overlook Ventures, Webb Investment Network
All funding rounds in the database
Product
AI that scores each P&C insurance policy for profit before and after binding
Insurance
Soteris is an AI platform for property and casualty insurers that predicts how much each policy will earn or lose, before the insurer writes or renews it. It is aimed at carriers whose books hide negative-profit policies and underpriced risks behind healthy averages, with one view for teams focused on EBITDA growth and another for teams focused on loss ratio. Soteris says it is purpose-built AI rather than an LLM, and that it requires no changes to rates, forms or filings. Scores are delivered through an API in under 250 ms, and the company says deployment takes about 90 days. It states that it uses no PII, holds SOC 2 Type 2 and has NIST AI RMF work underway. The company reports 100M+ submissions and $180B+ in premium scored.
Agent-Ready 0/5
- Discoverable — Can an agent find and read it?Not found on 3 November 2026
- Documented — Can it understand the API without a login?Not found on 3 November 2026
- Testable — Can it try the API without a sales call?Not found on 3 November 2026
- Connectable — Can it connect directly (MCP)?Not found on 3 November 2026
- Trusted — Can it see who runs it securely?Not found on 3 November 2026
Not in our registers
Standard v0.1.1 · checked 3 November 2026
Description from the company website · Source
In the news
- ICYMI fintech funding round-up: N.G CASH, Soteris, Benford, and more (opens in a new tab)
- Soteris launches AI tool to identify insurance policies that reduce profits (opens in a new tab)
- Soteris raises $8m seed to fix hidden P&C profit leaks (opens in a new tab)
Headlines from third-party publications, matched automatically. Links go to the original source.