Almetra Raises €16.3M Series A: How AI-Powered Video Analytics Is Making Factories 20% More Productive

Almetra, the Berlin-based manufacturing intelligence startup formerly known as Deltia, has closed a €16.3 million ($19 million) Series A funding round to expand its AI-powered platform that turns factory floor video into live operational data. The round was led by transatlantic growth investor blisce/, an early backer of Spotify and Pinterest . Here are the details every manufacturing and tech leader should know about the company, its technology, and its plans.

Funding Round: Who Invested and How Much

The Series A round closed in June 2026 and brings Almetra's total cumulative funding to approximately €20.8 million, following a €4.5 million seed round in April 2024 . The round was led by blisce/ and saw participation from six institutional co-investors: NAP, Merantix Capital, Robin Capital, Underline Ventures, Critical Ventures, and a syndicate of business angels .

Founded in 2022 out of the Berlin AI venture studio Merantix as Deltia, the company rebranded to Almetra in 2026 to reflect a broader ambition: moving from analyzing manual production steps to becoming the data and control layer for the entire factory . The team currently numbers about 40 people .

How the Platform Works: Three Data Layers Fused by AI

Almetra installs AI-powered cameras above factory production lines to capture video of manual and semi-automated processes, including assembly, packaging, and machinery adjustments . But the system isn't just a surveillance camera — it creates a unified model of production by combining three distinct data layers:

• AI video analysis — Computer vision algorithms analyze hand movements, cycle times, and operator actions in real time, converting raw footage into structured operational metrics such as throughput, downtime, utilization, and output rates .
• Machine data — The platform integrates with existing Manufacturing Execution Systems (MES), Programmable Logic Controllers (PLC), and Enterprise Resource Planning (ERP) software to layering equipment data on top of what the cameras see .
• Operator knowledge — The system incorporates human input to contextualize what the AI observes, helping distinguish between intentional process variations and genuine inefficiencies .

The result, according to the company, is a single real-time model of how work actually unfolds on the production floor. Teams can then identify bottlenecks and inefficiencies based on objective evidence rather than assumptions .

Real-World Results: 20% Productivity Gains at Bosch, Siemens Energy, and ABB

Almetra reports that factories using its platform achieve roughly 20% productivity improvements. The company's customer list already includes industrial heavyweights like Bosch, ABB, Siemens Energy, and Viessmann .

The productivity gains come from replacing manual observation and guesswork with continuous, objective data. Instead of relying on periodic time studies or operator reports, plant managers get a live dashboard showing exactly where time is being lost, which process steps are taking longer than expected, and where non-value-added activities are occurring .

Privacy First: How Almetra Avoids 'Big Brother' Concerns

Putting cameras on factory floors naturally raises worker privacy concerns. Almetra addresses this with a privacy-first architecture built on two principles:

• On-site video anonymization — Video footage is processed and anonymized locally at the factory. Identifiable imagery — faces, identifiable body features — never leaves the premises .
• No raw video leaves the facility — Only anonymized operational metrics, such as cycle times and utilization rates, are transmitted off-site. The raw video stays on the customer's infrastructure .

This approach is designed to comply with strict EU data protection requirements under GDPR while still allowing the AI to analyze detailed movement and process patterns needed for optimization .

What the €16.3M Will Fund: Product, US Expansion, and Robotics

Almetra plans to deploy the Series A capital across three strategic priorities:

• Product development — Accelerating the AI platform's capabilities, including deeper integration with factory automation systems and enhancing the proprietary large temporal models that analyze video on edge hardware .
• US expansion — Building a direct sales presence and customer base in the United States, a key target market where the company sees strong demand for shop-floor intelligence .
• Future robotics applications — The longer-term vision is to use the platform's deep understanding of human workflows not just to analyze processes but to automate them with robotics. By documenting exactly how work is currently done, the Almetra stack becomes an automation readiness and deployment engine, effectively telling robots what to do and where to do it .

The Bottom Line

Almetra's €16.3M Series A — led by Spotify/Pinterest backer blisce/ with six co-investors — validates a growing thesis in industrial technology: that computer vision, integrated with existing machine data and human knowledge, can deliver immediate productivity improvements while building a bridge to future factory automation. The company's privacy-conscious architecture and blue-chip customer roster give it a strong position as it targets the US market and looks to turn factory intelligence into factory action.