Meet Brian Remmington: M-Files Chief Software Architect Q&A

In May, Brian Remmington joined M-Files as Chief Software Architect, bringing more than 20 years of content management experience with him to lead the overall architecture and software design of the M-Files portfolio. Brian's background connects decades of document management experience with the next wave of enterprise AI.

We discussed the changes he's seeing with AI, the biggest challenges CIOs are facing, why context-first document management is essential, and his thoughts on what's next.

What are some of the biggest changes you've seen in the document management industry?

AI has caused a fundamental shift in the way organizations view their unstructured information assets. Previously, they stored documents to meet regulatory requirements and documents were viewed as a business cost. Now they are starting to understand that those documents hold huge business value.

Most organizations today have several document management systems of varying capability and quality, and while many can describe what good document management looks like, few have achieved it across their information estate. A lack of governance and quality assurance over documents stored in the various information silos is now thwarting effective AI use.

I joined M-Files because I saw that it wasn't the typical content management vendor. The platform does more than just store documents; it also manages business objects and their relationships to model people, products, organizations, processes, and information. Together, these objects and the web of relationships between them provide the curated business context needed for AI models to be trustworthy.

What is the biggest challenge you're seeing from CIOs right now?

I've been helping CIOs, CDOs, and Heads of Information Governance solve their information management problems for a long time, and what I'm hearing most is concern about poor context quality leading to untrustworthy AI outcomes.

Everyone wants to use AI to move faster and more efficiently, but if information quality and governance are missing, they're going to fail. Often, there's a mix of quality and outdated information in most document management systems and organizations are trying to curate information after the fact, which is an expensive and time-consuming approach.

What are your recommendations for organizations looking to get AI ready?

With any project, you need to focus on the problem you're trying to solve first. If you'd like to automate or accelerate an important decision, then identify what information a human would need to make that decision, then curate the right information that can be used to guide the AI. Start by solving one use case at a time, evaluating the output until performance is consistently strong. Then you can use what you learn to scale your program of improvements.

What are some of the other challenges you're hearing from peers?

People are increasingly comfortable with the idea of AI, but there's still a very significant trust gap. This often links back to the quality information feeding the model: if the context isn't accurate, governed, and curated, the output cannot be trusted.

We hear a lot about human-in-the-loop because people don't trust AI outputs and decisions. This signals that AI is still acting as an advisor rather than a trusted decision partner. Putting a human downstream of the AI addresses the symptom; improving the quality of the context upstream addresses the cause.

What changes do you expect to see in the near future?

The reinvigoration of context-rich document management will continue to build, but people will need to assess the format of their information, and determine which formats are best for AI to understand, interact with, and act on. For many years, document management providers have said it's about "getting the right information, to the right person, at the right time." In the AI-enabled future we'll need to factor in the reason the information is needed and what it's going to be used for so we can provide it in a format AI can use efficiently to make better, more trusted decisions.