Feature
Until recently, document management has focused on where files are located. But that’s changing as vendors are adding support for business context to knowledge work.
The problem starts with how we've always thought about documents, said Tony Grout, chief product and technology officer at M-Files. Traditional tools create operational friction by trapping knowledge in static folders and siloed applications, forcing people to remember where files live rather than why they matter. The alternative flips the paradigm.
"A context-first model changes the starting point to the business drivers, whether it's clients, projects, products or something else," Grout explained. "This transforms documents into dynamic knowledge by mapping connections between content and business information."
M-Files built its recently released Workspaces product on this principle: a system that mirrors business structure by mapping how people, projects and processes interact to get work done. Instead of navigating endless folders or relying on memory, employees start with what matters most — a client, project or process — and see the full picture in context.
"Workspaces represents a fundamental shift in how people work with documents and critical business information," Grout said. "M-Files is built on a model that understands how your business operates, delivering insights you can trust, automation that feels natural and a frictionless experience for a more human way to work."
The shift means work becomes available in the right business context through a knowledge graph that adapts to organizational needs. An account manager retrieves a statement of work through a client workspace, while a project manager views the document through a project lens: same file, different context, less confusion about versions or relevance.
The lead technologist at VASS, Michael Pytel, has spent his career in enterprise IT watching this problem play out.
"Traditional document management treats files as isolated objects, which means people waste time hunting for legacy information about the document like who created it, why decisions were made and what actions depend on it," Pytel said. Context-first models attach people, systems, decisions and workflows to documents, reducing version confusion and preventing information from getting trapped in personal folders.
Why Business Context Is What Makes AI Actually Work
The implications become more important when AI enters the picture.
Monika Malik, lead data and AI engineer at AT&T with prior experience at Barclays, focuses on production-grade systems that require accuracy. Current document management implementations have fundamental problems, she said.
"Folder entropy creates duplication, brittle paths and version sprawl," Malik noted. Moreover, for AI deployment, text-only systems operate without grounding; they process words but don't understand relationships between documents, people and processes.
Grout frames this challenge in terms of what AI needs to reliably function. Contextual AI uses metadata and business knowledge graphs to connect unstructured content with structured business data, grounding insights in business context rather than statistical pattern matching.
M-Files Workspaces works with Aino, an AI engine that draws on the context of clients, projects and processes rather than analyzing text in isolation.
"Without strong context, AI becomes noisy, generic and unreliable," Grout said. "With context, AI is significantly more accurate, hallucinates less and delivers answers that are precise, personalized and aligned to the user's intent."
"Folders describe where a file lives; context describes why it matters,” Malik said. “AI only scales on the latter."
From Folders to Knowledge Graphs
That scaling challenge requires architectural changes. Context-first management changes how AI systems operate, said Artur Borycki, vice president of AI research and development at Teradata.
"Most AI today operates on raw data — blind pattern matching," Borycki explained. "Context-first management changes this by embedding semantic relationships directly into the architecture."
The difference isn't just capturing that a customer purchased a product, but maintaining the web of information around customer role, industry pain points, trigger events, product capabilities and causal links. This knowledge graph helps change AI agents from correlation engines into reasoning systems that understand causation.
Borycki calls this simulation-enhanced analytics. Once rich context exists, AI stops looking backwards and starts looking forwards. Instead of probabilistic predictions based on historical patterns, organizations get strategic exploration across thousands of scenario variations with confidence intervals and sensitivity analysis.
"Strategy becomes continuous simulation rather than annual planning cycles," Borycki said. "Every decision comes with explored alternatives and probability distributions."
Speed Decision-Making With Shared File Context
The effect of faster decision-making is measurable, Borycki said.
Unified context layers reduce the coordination tax that slows most organizations, Pytel said. "When all contributors and their work sit inside one shared context layer, teams stop chasing missing comments or doubting which version is approved," he said. "Decisions move faster because every stakeholder sees the same history meaning the document is updated in real time."
Webinar
Dec
11
Fuel Your Front Line: Workforce Trends Shaping 2026
Get expert insight into the tools and tactics top frontline teams will need in 2026.
Register
Webinar
On demand
Fix the Disconnect Between Projects and Profit
Your margins are hiding in disconnected data. Let's find them.
Watch Now
Webinar
On demand
AI in Customer Service: Faster Resolutions, Happier Customers
Don’t let rising demand burn out your team. See how to build a smarter, more resilient support org.
Watch Now
Webinar
On demand
Accelerating Healthcare Ops with AI
Sign up to see how your peers are putting AI to work — and making it stick.
Watch Now
Webinar
On demand
From Nice-to-Have to Non-Negotiable: Prove the Value of Your EX Stack
Transform your tech stack: prove its value, secure your future.
Watch Now
Webinar
On demand
The Great Workforce Breakdown (And How to Stop It)
Disconnected employees. Failing communication. It’s time to turn things around.
Watch Now
Grout's team at M-Files documented how context-driven Workspaces reduce the time employees spend searching, backtracking, asking clarifying questions, second-guessing versions or jumping between systems. Instead, people get quick access to information needed for confident decisions because shared context and visibility into dependencies keeps teams aligned around a single source of information.
Workspaces delivers this through role-based, personalized views that find what each person needs most — including documents stored natively in Microsoft 365 — moving between related items without backtracking or re-filtering.
Context-driven systems are more auditable because every decision, system touchpoint and contributor gets logged, producing stronger evidence trails and fewer audit surprises for regulated industries, Pytel said.
Grout emphasized this dual benefit. The approach turns content into a system that captures, uses and governs information automatically, keeping it accurate, secure and audit-ready while reducing compliance risk through consistent classification, automatic permission enforcement and proactive gap identification.
How to Make Context-First Document Management Stick
Grout frames the adoption challenge around three factors:
- Capturing human knowledge and relationships around documents to create context.
- Automating processes and governance to reduce operational friction.
- Having documents contribute to a connected, trusted system of record.
Organizations need to recognize that documents aren't just costs to manage but assets that drive business performance, according to Grout. The shift requires working in familiar environments like Microsoft 365 while automation and metadata handle complexity behind the scenes, embedding context-first capabilities into existing workflows to avoid changing processes.
Malik, whose work focuses on policy-to-pipeline translation, emphasized the practical starting point. "Start with the business ontology: define 10 to 20 core entities and relations," she advises. "Don't over-model day one."
Malik’s approach includes metadata minimum viable products with small, mandatory fields that matter, such as owner, record type and customer, retention, plus policy and change management that aligns responsibility matrices and access models before rollout.
The platform itself needs to become the default place where work happens, not another destination to upload files, Pytel said. Clean integration with systems of record, sensible metadata design and governance that doesn't slow people down all contribute to success.
"Change management must encourage teams to shift from document-centric habits to transparent work," Pytel said. "Executive sponsorship matters too. Context-first models succeed when leaders reinforce that shared context drives business value."
Integrating Context With Enterprise Systems
That shared context often originates in core systems such as CRM and ERP platforms. Integration between these systems creates operational storylines rather than isolated transactions, Pytel said
"Integrating context-first environments with CRM and ERP systems turns operational data into a connected storyline instead of isolated transactions," Pytel explained. "Leaders can immediately see how a customer request, system issue or financial impact ties back to the underlying documents and decisions."
This requires technical architecture behind the scenes, Malik said. Reliable systems-of-record embed authoritative fields into workspaces, with bidirectional events where CRM or ERP changes update context and trigger approvals. This creates end-to-end visibility from opportunity through renewal with consistent policies across systems.
The result reduces silos while stakeholders see identical up-to-date information through views natural to their work, Grout said. Workspaces use business data from integrated systems to organize information, process workflows and apply appropriate access controls that mirror business operations. Deep integrations with platforms such as CRM, ERP and practice management systems deliver the context needed to drive productivity.
Context-First Models Turn Files Into Business Knowledge
Grout's paradigm flip redefines what documents represent in organizational work.
Traditional systems trap knowledge in static folders where information becomes hard to find and easy to lose. Context-first models organize around business drivers and dynamically map connections.
The organizational effect is that context layers capture reasoning, ownership and system signals, meaning institutional knowledge persists despite personnel changes, Pytel said. Teams avoid duplicating work or dealing with irrelevant documents because platforms find what matters to current tasks.
"By linking everything together, it prevents critical insights from disappearing into emails, desktops or individual memory," Pytel said. "This reduces operational risk and creates long-term resilience in knowledge-intensive environments."
Malik frames the technical implementation through production system requirements. Provenance-linked summaries let AI generate briefs with citations to canonical documents. Lifecycle management and ownership mean every artifact has an owner, purpose and expiration date, while deduplication by design clusters near-duplicates and flags canonical sources.
Grout sees this as removing dependence on institutional knowledge and personal filing habits so information doesn't vanish when employees change roles or leave. Context helps AI to guide users through vast information stores, reducing overload by making knowledge easier to consume.
The approach means knowledge stays discoverable and presented in appropriate context, reducing risks from missed insights while preventing overload by finding only what's relevant to tasks at hand. As organizations navigate increasing complexity and AI becomes more central to operations, that shift from managing files to managing context may determine which enterprises thrive and which drown in their own information.
Editor's Note: Catch up on more coverage of how the information management space is evolving below:
- Enterprise Content Management Gets Disrupted, Again — Keeping up with all the name changes ECM has gone through over the years is hard. But the latest – intelligent content services – might just have holding power.
- What's Next for Intelligent Document Processing? — Intelligent document processing (IDP) continues to evolve beyond document capture to become a key part of enterprise transformation.
- Why Hyland's New CEO Says 'ECM Expertise Is One Thing We Don't Need' — A new leadership team, a future vision focused on content intelligence, a customer base hungry for innovation: Hyland is primed for its great pivot.