velou

Velou provides product data infrastructure for ecommerce retailers, analyzing raw product text and imagery to generate missing attributes, standardized tags, and richer content, and to build product graphs that map relationships between items. The company positions this enriched data as the layer AI shopping and search agents need in order to reason about products, curate alternatives, and recommend compatible items. Enriched data, tags, and structured attributes are pushed back into a retailer's existing commerce platforms rather than replacing them.

The core technology is Commerce-1, a model the company describes as built exclusively for retail and trained on 4 trillion tokens of retail-only text and imagery, then refined with reinforcement learning from merchandisers. The architecture uses dedicated attention heads for attributes such as color, material, fit, and seasonality to generate structured attributes and variant groupings, and layers price-integrity checks, brand-tone conditioning, and regulatory filters on outputs. The model is described as retraining nightly on fresh click signals.

Velou also exposes a REST API for product indexing and search. Documented capabilities include single and bulk product indexing (up to 500 products per request), product deletion, task tracking, search results and query suggestions, similarity, "Complete the Look" recommendations, event-tracking analytics, and content-generation webhooks. Access uses key-based authentication over SSL with separate admin, search, and analytics API key types.

Founding story

Velou was founded by Sadee Gamhewa, who previously led product, innovation, machine learning and AI strategy, and engineering at multiple startups and large corporations including Fortune 100 companies, holds patents in ecommerce, and has published machine learning and AI research.

Business model

Velou sells software to retailers and ecommerce businesses, integrating with their existing commerce platforms and delivering enriched product data via APIs and platform apps, including a Salesforce Commerce Cloud app. Commerce-1 usage is metered in tokens, with existing customers offered 1 million free tokens and new users able to join an early-access wait-list.

Software sold to retailers, with Commerce-1 consumption measured in tokens.

Traction

The company states Commerce-1 is trusted across more than 100 million products and cites customer case studies including growth for a furniture marketplace, a 3x conversion increase with AI smart product catalogs for books, and reduced customer frustration and increased revenue at Get The Label. A third-party company directory lists Velou at 11-50 employees.

Latest developments

Velou announced Commerce-1, described as its largest launch to date, with availability to existing customers and a wait-list for new users; a launch video features CEO Sadee, lead ML engineer Jayana, and customer-success lead Debby. Developer documentation records a version 2.0 API release on February 17, 2026 adding a v2/products endpoint with updated validations, following a 1.9 release on October 10, 2025.