Cognicon's keynote speaker was Chris Hill, Senior Director at Pragmatic Semiconductors, who gave a fascinating talk about the evolution of digital governance throughout his professional life.
He told the story of his journey from Project Manager with the British Antarctic Survey to more recent roles with fast-growing medical device and semiconductor firms. He reflected on how generations of digital tools have helped (and sometimes hindered) the pursuit of quality and competitive edge in some of the world's most complex industries.
With this context in mind, CogniCon kicked off.
Following Chris Hill's keynote, Cognidox's CTO Víttal Aithal shared exclusive product and roadmap updates about the Cognidox platform:
As the day continued, Simon Roberts from Cognidox hosted two panel discussions:
The first discussion, headlined "Product Development: Product Realisation, the Journey from Concept to Market.", featured a stellar line-up:
Kevin Newley talked about how document control helps strike a balance between agility and governance in the development of Graphcore's IPU-based AI systems. Rupert Baines and Peter Claydon spoke about the structuring role that Cognidox played in the product development process that led to the multi-million dollar acquisitions of their semiconductor startups.
"Cognidox was introduced as a mandate to our teams at the start, but it grew organically throughout the company as users began to realise it could solve more problems."
Rupert Baines
The second discussion focused on medical devices. The expert panel comprised:
This session looked at the challenges of compliance and quality management in medical device development, particularly within start-ups and scaling companies. Jonathan Wyn-Beer talked about his time at Coalesce and how the complexity of documenting user requirements and conducting design reviews was eased through the use of Cognidox.
Rachell Rodell outlined her work at UCL, helping academics use a structured process to turn medical device concepts into compliant products. Sam Shelley outlined the importance of medical device start-ups building a flexible eQMS from the ground up based on document control - rather than imposing compliance with heavy-duty software systems.
‘You don’t drop a QMS onto a business like a fire blanket on a chip-pan fire.”
Sam Shelley, SQMS
In the final talk of the day, Peter Brady from consultants UOVO Medtech gave a much-anticipated presentation on the future of AI in medical device development.
This talk explored AI's transformative potential in enhancing efficiency and quality across the medical device software lifecycle while helping navigate complex regulations. He drew parallels between the current AI hype and the dot-com bubble, suggesting that while a correction may occur, the underlying technology will be truly transformative.
As much of the regulations are still being written in this area, there is still uncertainty about what developers will be required to do and how AI can be used in the future.
“Machine learning validation is over-reliant on testing. Conventional safety-critical software development has a comprehensive validation toolkit that goes beyond testing. Machine learning needs to develop similar techniques. This is an active area of research for us”
Peter Brady, UOVO
Peter spoke about the challenges in validating and testing AI/machine learning systems for safety and efficacy, as well as the powerful opportunity the technology offers to accelerate and derisk development and streamline compliance.
The talk touched on UOVO Medtech's AI-driven requirements management tool, which will deliver a single source of truth for medical device developers. The tool will automatically generate Word documents that will be added to Cognidox and routed for signatures. It will analyse requirements for quality, suggest rewordings, generate additional requirements, automatically create Jira tickets, and ensure synchronisation between documentation and code implementation. UOVO is also building APIs in Cognidox to support the EU AI Act's data logging requirements for high-risk AI systems.
CogniCon proved a rare chance for our customers and industry friends to hear from sector leaders about the synergies between document control and quality management in the life science and high-tech industries.
Guests were able to meet the Cognidox team, network with other customers and look to the future with their peers and industry experts.
We look forward to bringing you more of the same at CogniCon 2025!