Convr vs Gradient AI — Underwriting workstation for US insurance, 2026.
Convr (7 named carriers) and Gradient AI (8 named carriers) both sit at the underwriting workstation layer. Zero customer overlap in the public roster — they are addressing different segments of the same stack layer.
Last verified 2026-05-29 · methodology
TL;DR
- Convr has 7 publicly-named carrier deployments; Gradient AI has 8. Both at the underwriting workstation layer.
- Zero customer overlap in the public roster. Convr and Gradient AI are addressing different carriers within the same stack layer.
- Both classified ai-native on Phidea's generation axis.
- Both independent ownership.
- Analyst coverage: 0 firms cover both, 3 only Convr, 3 only Gradient AI.
Customer overlap
| Bucket | Count |
|---|---|
| Named on Convr only | 7 |
| Named on Gradient AI only | 8 |
| Named on both | 0 |
| of which US-named on at least one side | 0 |
Only on Convr
- Zurich North America (US)
- Hiscox USA (US)
- Penn National Insurance (US)
- Encova Insurance (US)
- Selective Insurance (US)
- WCF Insurance (US)
- Columbia Insurance Group (US)
Only on Gradient AI
- The Builders Group (US)
- AmFed (US)
- BTIS (Builders & Tradesmen's Insurance Services) (US)
- Signal Mutual Indemnity Association (US)
- Allied National (US)
- ATS Underwriting (US)
- Skyward Specialty Insurance (US)
- North Carolina League of Municipalities (US)
Counts derived from 15sourced carrier-deployment entries across both vendor cards. Aggregate-only statements (e.g. “16 of the top 20”) excluded.
Stack position
- Generation
- ai-native
- Stack layer
- Underwriting workstation
- Founded
- 2016
- Lines
- commercial, specialty
- Generation
- ai-native
- Stack layer
- Underwriting workstation
- Founded
- 2018
- Lines
- workers-comp, health, commercial
Ownership and corporate context
Carrier-segment specialization
Convr — geographic split
- US7
Gradient AI — geographic split
- US8
Analyst coverage differential
Only Convr cited by
- Carrier Management (2020: DataCubes Becomes Convr, and Its CEO Explains Why)
- Reinsurance News (2025: Zurich North America enhances underwriting efficiency with Convr AI)
- PR Newswire (2025: Convr AI Holds the Universe of Commercial Insurance within Submission Ontology)
Only Gradient AI cited by
- Digital Insurance (2018: Gradient A.I., spun out of Milliman, looks to midsize insurers for growth)
- SiliconANGLE (2024: Gradient AI secures $56M to enhance insurance industry efficiency)
- InsurTech Digital (2023: Signal Mutual Integrates Gradient AI for Claims Management)
Recent news (last 12 months)
No news items in the last 12 months for either tool.
Sourced limitations
- Convr's total disclosed funding is approximately $18M across two rounds, ending with the $15.2M Series B in November 2019. No later equity rounds have been publicly announced. That capital base is modest compared to well-funded underwriting automation peers, which may constrain the pace of model development and international expansion.Source: PR Newswire
- Convr is a US commercial P&C specialist. Its submission ontology and carrier integrations are built for the US market — standard and specialty commercial lines. It does not serve personal lines, life, health, or workers' comp as primary lines, and no European or Asia-Pacific carrier deployments have been publicly announced.Source: Convr
- Convr's accuracy claim for machine-read data is 91%, which means roughly 1 in 11 data points still requires human review or correction. For high-volume small commercial portfolios this is workable; for complex middle-market risks with long loss-run histories, residual manual touches remain.Source: Convr
- Convr does not replace a policy administration system or rating engine — it sits upstream of those. Carriers that lack mature APIs into their policy admin stack will need integration work to route the structured output Convr produces into downstream underwriting decisions.Source: Convr
- Gradient AI is not an underwriting workstation or a policy admin system. It ships risk scores, loss-ratio predictions, and claim-triage signals that carriers and MGUs consume via API or embed into existing underwriting and claims workflows (e.g. Origami Risk for The Builders Group, Duck Creek via a named partnership). Replacing a PAS or a claims admin system is out of scope.Source: Duck Creek Technologies
- Despite marketing references to 'all major lines of insurance', Gradient AI's productised coverage is concentrated in workers' compensation, group health (including medical stop-loss) and general/commercial P&C. No dedicated life insurance product surfaced in press releases or product pages reviewed here; MassMutual Ventures is a minority investor, not a life underwriting customer.Source: Gradient AI
- No Gartner, Forrester or Celent dedicated vendor profile on Gradient AI surfaced in public search. Third-party coverage is concentrated in trade press (Digital Insurance, SiliconANGLE, InsurTech Digital, Insurance Business America) and Gradient's own Business Wire releases — buyers relying on analyst rankings will find the signal thin.Source: Crunchbase
Limitations published on Phidea are sourced to the underlying citation and reflect what is publicly named — not an exhaustive list. Consult the vendor card for the full record.
Frequently asked
- Do any carriers run both Convr and Gradient AI?
- Not in Phidea's public roster. Across 15 sourced carrier-deployment entries on both vendor cards, zero carriers appear on both. The two tools are addressing different carriers within the same stack layer.
- Who owns Convr and Gradient AI?
- Convr is independently held. Gradient AI is independently held.
- Which has more named US carriers?
- Gradient AI has the larger publicly-named US roster: Convr 7, Gradient AI 8. Public-roster size is a coverage signal, not a quality signal — vendors with stronger NDAs may have larger actual US footprints than the public count shows.
- Where are these tools positioned in the insurance stack?
- Both sit at the underwriting workstation layer. Convr operates as a standalone vendor; Gradient AI operates as a standalone vendor.