Polysense raises €9.4M oversubscribed Seed to automate AI quality control across food manufacturing

Polysense, a Ghent-based AI company, has raised an oversubscribed €9.4 million ($10.7 million) Seed round led by Felix Capital, with participation from Fortino Ventures. The capital goes to expanding the company's AI quality control platform across more stages of the food production process and to funding global rollout.

Automating the production line inspection problem

Quality control in food manufacturing is an operation that runs at the intersection of food safety regulation, production speed, and labour cost. Every production line requires inspection — for defects, contamination, non-conformities, and compliance failures — but manual inspection is slow, inconsistent, and does not scale with production volume.

Polysense applies AI and computer vision to automate that inspection layer. The platform analyses production in real time, detecting defects and non-conformities across the food manufacturing process. The model is trained on food-specific failure patterns — the types of errors that appear in food production are different from generic manufacturing defects and require specialised training data to detect reliably.

The company's pitch to food manufacturers is twofold: inspection that is faster and more consistent than manual processes, and coverage across multiple stages of the production process rather than just the final checkpoint. A defect caught early costs less to address than one caught at the end of the line — or, worse, by a retailer or consumer.

Felix Capital's thesis

Felix Capital is not a food tech specialist. The London-based fund is known for consumer and lifestyle technology — its portfolio includes Farfetch, Flink, and Spotify. Its decision to lead a Seed round in a food manufacturing AI company signals a thesis about enterprise software scale in a large, traditionally under-automated market, rather than a sector rotation.

The oversubscribed close at Seed, with Fortino Ventures — a Belgium and Netherlands-focused B2B technology investor — adding geographic and sector expertise, indicates the round attracted real competitive interest.

The expansion problem

The global food manufacturing market is large and heterogeneous. A quality control platform that works for one category of food — fresh produce, packaged goods, bakery products — does not transfer automatically to another. Training data is specific, production environments vary, and regulatory requirements differ by product type and market.

The €9.4 million goes to expanding coverage across more production stages and food categories, and to the international rollout that moves Polysense from a regional player to a global platform. The 18-month test is how many categories and geographies the platform can credibly cover before competitors — including both specialised startups and larger machine vision incumbents — respond to the same market opportunity.