Tractable
Computer-vision AI that turns smartphone photos of vehicle or property damage into line-level repair estimates and claims triage in seconds.
tractable.ai ↗Score
- Traction (named carrier deployments)8 carrier deployment(s) with public source.
- 3/5
- Maturity (years since founding)12 years since founding (2014).
- 4/5
- Coverage (insurance lines supported)2 line(s) supported: auto, home.
- 2/5
- Analyst recognition (Celent / Gartner / Forrester / Everest / ISG)2 mention(s).
- 2/5
What it does
Tractable is an AI-native damage-appraisal vendor founded in London in 2014. Its core product ingests smartphone photos or video of a damaged vehicle or property and returns a line-level repair estimate plus a triage recommendation.
Business footprint. Tractable reached unicorn status in June 2021 with a $60M Series D co-led by Insight Partners and Georgian, bringing total funding to $115M at that point. It followed up with a $65M Series E in 2022, taking cumulative funding to approximately $180M. TechCrunch reported at the 2021 round that Tractable was serving 20+ of the world's top-100 auto insurers, including Covéa (France) and Ageas (UK).
Public carrier deployments. Admiral Seguros (Spain) and Beesafe (Poland) are cited in Tractable's own case-study materials. In the Admiral case, 70-75% of policyholders who receive the AI web-app link complete their claim digitally in roughly two minutes, and AI review of an estimate happens in seconds versus ~30 minutes of adjuster time.
What it replaces. The dominant legacy workflow is adjuster photo review inside CCC ONE or Mitchell. Tractable sits either upstream (triage) or parallel (estimate generation) to that workflow.
Analyst gap. Despite the scale, Tractable has not shown up in a major Gartner/Celent/Forrester leader quadrant for insurance AI in publicly indexed 2023-2025 reports. Coverage is concentrated in trade press (InsurTech Digital, Repairer Driven News).
What it does not do. Per Repairer Driven News' 2024 analysis, Tractable is positioned as augmentation for claims handlers — edge cases, fraud signals, and complex liability still route to humans.
Common questions
What does Tractable actually do? Tractable takes photos or video of a damaged vehicle or property and produces a line-level repair estimate in seconds. It also gives a triage recommendation so a claims team knows which files need a human adjuster and which can be settled automatically. The output plugs into existing claims platforms rather than replacing them.
Which carriers use Tractable? Publicly named customers include Admiral Seguros (Spain), Beesafe (Poland), Covéa (France), Ageas (UK), GEICO, The Hartford, Root Insurance, and American Family Insurance. TechCrunch reported in 2021 that more than 20 of the world's top-100 auto insurers were using the platform.
How does Tractable's photo-based estimating differ from a human adjuster? A human adjuster reviewing photos in a legacy platform like CCC ONE or Mitchell typically takes around 30 minutes per estimate. Tractable's model processes the same photos in seconds and flags whether the damage matches the claimed scenario. The adjuster still handles the final decision on edge cases, fraud signals, and complex liability.
Is Tractable used outside auto claims? Yes. Tractable has a property damage product that applies the same computer-vision approach to home claims — roof, siding, and structural damage from photos. It lists home alongside auto as a supported line. The property product is newer and less documented in public case studies than the auto product.
Named deployments
- Admiral Seguros (ES)Tractable
- Beesafe (PL)Tractable
- Covéa (FR)TechCrunch
- Ageas (UK)TechCrunch
- GEICO (US)tractable.ai
- The Hartford (US)tractable.ai
- Root Insurance (US)tractable.ai
- American Family Insurance (US)tractable.ai
Known limitations
- Tractable's AI still requires human adjuster oversight for edge cases and is positioned as a triage and estimation tool, not a full replacement for claims handlers — per trade-press analysis of its capabilities in 2024. (Repairer Driven News)