Ameba

Ameba (legal entity Ameba Systems Ltd.) is a London-based software company that sells what it markets as an "AI workforce" for supply chain administration. Customers connect the platform to the ERP, email inbox and messaging tools they already use, then describe their standard operating procedures in plain language — what to watch for, what action to take, and when to escalate to a human. Ameba then applies those processes to every order continuously, reading incoming emails, files and ERP updates and acting when a trigger matches: chasing suppliers, drafting quotes and replies, confirming dates, extracting document fields, matching invoices against purchase orders, and writing results back into internal systems. The company positions the setup as non-invasive, requiring no data migration.

At the time of its 2024 seed round the product was described as a generative-AI layer over existing supply chain software that extracts unstructured data from emails, PDFs, WhatsApp and Excel, surfaces it centrally, and predicts disruptions and bottlenecks for retailers. Public positioning has since shifted toward agentic execution of order-level administrative workflows for manufacturers, with the company describing itself as an AI teammate that keeps every order in sync with what suppliers and customers actually said.

Ameba states it is SOC 2 Type II and GDPR compliant. Its own site cites cumulative usage of roughly 50,000 orders and 80,000 line items tracked, 580,000 emails and 30,000 files processed, and 27,000 exceptions identified.

Founding story

Ameba was founded in 2023 by Cedrik Hoffmann and Craig Massie and is based in London. Hoffmann was previously supply chain director and co-founder of e-commerce company VALOREO; Massie is a former Palantir engineer. The company describes itself as built by former supply chain operators.

Business model

B2B software sold to manufacturers and retailers, delivered as a platform that integrates with the customer's existing ERP, email and messaging stack; onboarding involves configuring the customer's own processes, after which the agents run against every order. The sources do not disclose pricing or contract structure.

Traction

Company-reported cumulative usage: 50,000+ orders and 80,000 line items tracked, 580,000 emails and 30,000 files processed, and 27,000 exceptions identified. Claimed outcomes include a 10% increase in on-time delivery, up to 80% less manual data entry, and average delay cut in half; at the 2024 seed announcement the claim was a 30% reduction in manual data input. Anglotex's Head of Operations, João Gomes, reports raw materials arriving nearly 10% faster. Plank generated 140 alerts on critical production and delivery delays via Ameba.

Latest developments

The website markets an agentic "AI workforce" product with a one-time, roughly 10-minute workflow setup and reports cumulative processing volumes across orders, emails and files. The careers page lists no open roles at present.