Tractable

Every year, hundreds of millions of insurance claims are filed for physical vehicle damage worldwide. In almost every case, the process has remained unchanged for decades: a human adjuster inspects the vehicle, estimates the damage, files a report, and the settlement arrives days or weeks later. Tractable set out to attack this bottleneck with the same technology that powers self-driving cars: computer vision and deep learning applied to the visual assessment of physical damage. Tractable's platform analyzes photographs of a damaged vehicle — taken with an ordinary smartphone — and produces an accurate repair cost estimate within seconds. The system was trained on millions of real claims, enabling it to identify damage types, affected parts, and associated costs with accuracy that matches or exceeds that of an experienced adjuster, but at a radically different speed. Claim settlement time drops from days to minutes. Tractable's customers are insurance companies and auto dealers, both of which face enormous operational costs in managing appraisals and both of which benefit directly from automation in terms of efficiency and process standardization. For an insurer processing millions of claims per year, automating even a fraction of visual damage assessment translates into substantial operational savings. Founded in London by Alex Dalyac, Adrien Cohen, and Razvan Ranca, Tractable built its initial position around a specific use case — vehicle damage — before expanding the technology to adjacent domains such as natural disaster damage assessment. Its business model is built on SaaS licenses tied to the volume of assessments processed, generating recurring revenue aligned with actual system usage. The market for AI-driven claims automation within insurtech remains large and relatively underpenetrated, placing Tractable in a sector with meaningful entry barriers — training dataset quality, integration with legacy insurance systems, accuracy across an enormous variety of vehicles and damage types — and strong potential for global geographic expansion.