Epinium

Epinium develops an AI software platform for brands, manufacturers and marketing agencies selling through Amazon and other digital commerce channels. The product connects and manages multiple Amazon accounts from a single interface, integrating Seller Central, Vendor Central, Brand Analytics and Amazon Ads, and adds a Shopify connection plus other marketing channels. Once connected, users get account analytics, an audit that identifies actions to improve results, and optimization tooling for listing SEO and advertising built on the company's own AI models alongside integrations with third-party large language models.

The current positioning centres on "agentic AI": a set of agents that execute recurring commerce work — reporting, catalog updates, bid and budget adjustments, stock checks — under an explicit approval model the company calls a delegation ladder (agent proposes and the user approves; trusted domains run autonomously; governance through action logs, approval rules, spend limits and rollback). The knowledge layer, marketed as the "CMO AI", packages 100 marketing playbooks across 22 disciplines with 216 reference documents, each playbook carrying a formula, validity threshold and caveats; playbook categories span paid and retail media, marketplace and catalog, brand and content, demand and conversion, SEO/GEO and analytics, and direction and sales. Named components include Velax, an AI strategist; an MCP integration that exposes the same playbooks and account data inside Claude or ChatGPT; workflow and monitoring templates; and modules for advertising, catalog and content, and stock and operations. The company states the CMO AI can reach customer data through five paths in order of preference: deep Amazon integration backed by nine years of stored history, live light connections such as Shopify, whatever the user's own assistant already connects to, browser control of tools without connectors, and manual files, with 62 of the 100 playbooks usable from uploaded data alone. Alongside the platform, Epinium sells AI consulting and custom AI development deployed on the client's own infrastructure.

Trust and compliance are presented as product attributes: EU AI Act compliance, centralized and reversible logging of AI actions, channel credentials held in a separate encrypted vault with scoped revocable tokens rather than Amazon credentials exposed to assistants, and an Amazon data-security assessment completed in December 2025.

Founding story

Carlos Martínez and Miquel Ferrando founded a boutique full-service digital agency in 2015 offering branding, e-commerce platforms, mobile apps, SEO, data analytics and custom software development. In November 2017 they launched Roicos as an Amazon-specialized agency, and in November 2021 spun out Epinium as a niche SaaS product for listing management and automated SEO audits.

Business model

Epinium sells subscription access to its SaaS platform, offered with a seven-day trial, while the CMO AI playbook layer (including its MCP access from Claude or ChatGPT) is offered free of charge with no card required. A services line provides AI consulting and custom AI development built on the client's stack and deployed on the client's infrastructure with a fixed timeline and written guarantee, with the resulting work owned by the client.

Recurring software subscriptions for the platform, with a free tier for the CMO AI and a paid-trial entry point, supplemented by project-based AI development and consulting fees.

Traction

Epinium's site cites more than 500 brands using the product, while its about page reports 1,200+ active sellers, €300M in total sales generated, €20M in managed ad spend and 67 countries covered. At the time of the 2023 funding round the company reported more than 70 clients. Third-party listings place headcount at 11-20 employees. The company's stated vision is to serve more than 2,000 active clients in 20 countries by 2027.

Latest developments

Recent activity includes the 2025 rebrand and platform overhaul, the rollout of the free CMO AI with 100 playbooks and an MCP integration for Claude and ChatGPT, an Amazon data-security assessment completed in December 2025, and the launch of an AI consulting and custom development line. The company also publishes industry analysis on its blog, including coverage of enterprise MCP governance drawn from 12 brand-side MCP deployments it reviewed in Q4 2025 and Q1 2026.