
As the AI industry shifts to adjust to various regulatory and compliance certification demands, Deltia.ai is working with CertifAI to make sure its AI-based products adhere to the highest development standards.
Cross-Site Benchmarking
Quality Risk Detection
Bottleneck Detection
Building trustworthy AI for manufacturing compliance
As the AI industry shifts to adjust to various regulatory and compliance certification demands, there are some burning questions remaining: What does technical compliance actually entail? How do requirements vary across regions and industries?
We at Almetra have been asking ourselves similar questions. In the shifting landscape of the manual manufacturing industry, one thing is certain: Those who don’t adapt and evolve get left behind.
Computer vision, AI, and compliance
Lack of digitization in manual processes inhibits speed, better training, and overall improvement. Things fall through the cracks due to poor data and limited data gathering opportunities.
Almetra’s AI-driven computer vision models are unique within the market, and Almetra provides a depth of visibility of the shop floor that has not existed until now. Computer vision enables the collection of not just data, but also in-depth analysis of thousands of data points. Almetra easily identifies patterns, and suggests areas for optimization and improvement.
However, that means we as a company also fall within the scope of the EU’s AI Act, ratified in May 2024, with some of the new rules already coming into full effect late 2024. Full compliance will be a necessity for most high risk use cases in 2026.
We strive to make sure our product is always compliant, and the security of our customers and those working on the shop floor has always been one of our top priorities. That’s why we’re working together with CertifAI to make sure our Almetra’s AI-driven product meets the highest standards.
The AI Act and the demand for external testing
The full running length of the AI Act is hundreds of pages. 800 plus pages. To average it out, that’s comparable to the length of Leo Tolstoy’s Anna Karenina. If you prefer suffering for the sake of compliance, we’d recommend reading the AI Act in full. However, if you’d prefer to save your time and spend it instead on reading about the suffering of Tolstoy’s characters, then CertifAI is a much simpler solution.
Companies need to be prepared for the full scope of incoming regulation, including the AI Act and a mix of national and European sectoral legislation. CertifAI, the official AI testing and certification joint venture between Dekra, PwC, and the City of Hamburg, provides unique insights on regulatory environments and combines this with in-depth technical knowledge, including software-based testing for AI systems.
CertifAI is currently supporting Almetra in making sure our AI-based products adhere to the highest AI development practices, and mitigating potential risks in a structured way.
Finally, certification is not only a requirement, but also a competitive advantage. Through CertifAI’s unique testing methods, we can remain confident that we’re operating at the highest level of state of the art AI system development. Technical robustness is therefore a solid foundation for regulatory compliance — something that many companies in the sector can’t say.
Fostering innovation in AI
CertifAI is led by an experienced team of AI pioneers and legal experts that not only have a deep knowledge of the ever expanding regulatory framework and technical capabilities around testing AI, but also remain committed to making sure that innovation does not cease.
“Adherence to AI development standards is not only a horizontal and sectoral regulatory obligation, but should also be part of every company's software quality assurance. Testing and certification of complex applications play a central role in QA processes.” – Robert Kilian, CEO, CertifAI
We’re not just using CertifAI on our finished product — we’re also making sure we’re integrating AI best practices into our development process.
While the Almetra product enables access to context-rich shop-floor data via state of the art computer vision, we as a company never compromise on our privacy principles. Almetra anonymizes all video footage for the security and comfort of workers on the floor. This means that customers get unprecedented insight into their manual processes, without wading into the murky waters of GDPR and AI Act compliance issues.
At Almetra, we’re excited to work with CertifAI to make sure that all of our customers can feel confident in the safety and quality of our AI product. We see our collaboration as a mark of our commitment to excellence, and hope that it sets a standard for the industry.
Learn more about AI compliance and manufacturing AI
If you’d like to learn more about how Almetra is helping streamline manual manufacturing processes with AI and computer vision, you can go here.
To see how CertifAI can help you with AI compliance, testing and certification, check them out here.
Keep reading
Raw video deleted within 10 minutes
Only anonymised snippets and aggregated statistics are retained.
Edge processing on-site
No raw footage leaves the factory floor. Full cybersecurity compliance.
Works councils approved
Approved by works councils across Germany, Poland, and other co-determined manufacturing environments.
Privacy by Design
Work-council approved.
Edge-processed.
Privacy-preserving from the ground up.
Questions
answered.
What is Almetra and how is it different from a camera or video analytics system?
What kinds of factories and production environments does Almetra work in?
How long does deployment take — and what does implementation look like?
How does Almetra handle data privacy and worker concerns?
Does Almetra replace our existing MES, ERP, or PLC systems?
What outcomes can we realistically expect — and how quickly?
Is Almetra only useful for improving existing lines — or can it help with new products and ramp-ups?
What does the path from pilot to full deployment look like?
System status · OperationaL
v 2026.5
Watch recent customer story







