Elara Spatial

Elara Spatial, Inc. (operating at hosta.ai and elaraspatial.ai) builds a computer-vision platform that converts a small number of interior photographs taken on an ordinary smartphone into structured property data and a finished damage estimate for insurance claims. The workflow has three stages: a link opens the phone's browser camera so no site visit, app download, or specialized hardware is required; the system reconstructs the room in 3D so dimensions are derived from the model rather than measured by hand; and the platform returns a preliminary estimate in which each line item traces back to the source image and the carrier pricing rule applied to it.

The platform identifies building materials and assesses damage severity from imagery, applies carrier-specific assessment guidelines and workflows, and is designed to fit existing claims processes rather than require operational change. Personally identifiable information is stripped before processing, and the company states it uses computer vision rather than generative chatbots. Every estimate passes human-in-the-loop quality assurance, reviewed by Elara's team and a licensed adjuster before finalization; the company positions the output as objective documentation to support adjusters rather than as a claims decision. Integration is offered through APIs.

The company is based in Cambridge, Massachusetts, and is led by CEO Ingrid Flores and CTO Mica Arie-Nachimson.

Business model

Enterprise software sold to property insurance carriers, combining automated computer-vision processing with a human review layer staffed by Elara's team and licensed adjusters. Pricing is listed as custom or hybrid, negotiated with the provider rather than published, and the platform is delivered through API integrations into carriers' existing claims workflows.

Custom or hybrid pricing set through direct contact with the company; pricing is described as varying with usage volume and selected features, with custom enterprise arrangements available.

Traction

Reported production results include a 40%+ reduction in claim cycle time from first notice of loss to first payment, 50% more claims processed per adjuster without added headcount, over 80% of estimates returned without errors, and estimate turnaround of about two hours against a 28-day baseline. A third-party listing cites adjuster throughput gains of up to 125%. In September 2026 the company described multiple active carrier pilots and a growing pilot pipeline.

Latest developments

As of September 2026 the company reported reaching the performance standard carriers require for adoption, with multiple pilots running and an expanding pilot pipeline, and stated its goal of enabling carriers to tell policyholders what they are owed within two hours rather than roughly 30 days. The company's privacy policy was last updated on 25 September 2026.