Transforming KM as a catalyst for AI success with M-Files and Market Logic Software

AI is dependent on well-curated, fresh, contextual information that is both accurate and easily retrievable. With a robust KM foundation in place, fueling AI with this sort of information is possible—and necessary to unlock AI’s full potential.

Frank Taliano, senior director, product management at M-Files, and Joseph Rini, director product management at Market Logic Software, joined KMWorld’s webinar, Building a KM Foundation for AI to evaluate best practices and strategies for laying the KM groundwork crucial toward enabling the effective, responsible usage of AI.

Everyone is becoming well-accustomed to the various promises of generative AI (GenAI)—from driving creative innovation to automating menial tasks, providing personalized assistance, transforming knowledge access, and more. Though like its promises, we are also becoming acquainted with GenAI’s litany of challenges, such as lack of rich context, information security risks, and low-quality content inputs and outputs.

Fortunately, according to Taliano, “With the right information management system in place…there is a clear path to unlocking the potential of generative AI for knowledge management use cases.”

This system consists of two main building blocks, explained Taliano: curation and context.

Curation ensures that AI only accesses relevant and up-to-date information to ensure accuracy of responses for any given topic or task. Curation also emplaces granular access controls that enable AI to comply with an organization’s information security policy, preventing information leakage to unauthorized users. Ultimately, curation establishes credibility and trust through consistent, relevant, and useful responses, while simultaneously optimizing retrieval efficiency.

Taliano offered the following best practices for content curation:

• Establish a clear governance framework.
• Organize content with consistent metadata and business taxonomies.
• Maintain content accuracy and relevance through regular reviewing and automation.
• Enable version control and audit trails to keep track of all content changes.
• Implement automated access control based on user roles, business context, and metadata.
• Automate where possible, from tagging to validation and recommendations.

Context is the key to understanding, noted Taliano, enabling AI to respond to a user’s query correctly according to meaning and intent. M-Files Metadata structure creates a unique, customer-specific knowledge graph consisting of all content, fueling AI with the necessary information—and metadata—to deliver rich, contextual outputs.

“Metadata is what makes documents more than just files,” said Taliano. “It gives documents meaning and structure and makes them easier to structure and process. It goes well beyond simple organization; metadata is really a powerful tool that can contextualize information.”

Following Taliano, Rini examined a fundamental obstacle plaguing modern KM: The rapid rate of change and increase in knowledge overwhelms our ability to make decisions.

Yet, insights based on dynamic knowledge remain a crucial differentiator between business success and failure. According to Forrester, insight-driven businesses are 2.8x more likely to experience double-digit year-over-year growth, further supported by Greenbook’s findings that 80% of success factors in innovation are insight-related.

Powering an insights-driven, business success is GenAI, offering specialized capabilities that can adequately adapt to today’s chaotic world of KM. With Market Logic Software’s DeepSights solution, enterprises can bring trusted insights into daily work effortlessly, using retrieval-augmented generation (RAG) and GenAI to power KM.

DeepSights is specially trained to understand market research and competitor intelligence, explained Rini, compiling six layers of analysis to automatically understand content and extract relevant information.

“We have the benefit that content that ends up in our system is already the final draft…in coaching our users how to get the content into the system…in its final form…our RAG flow can really focus on identifying the most important documents in this huge subset of content,” said Rini.

Market Logic Software heavily emphasizes the importance of quality for its DeepSights solution, offering capabilities that enrich the evidence chunks extracted from market research intelligence. With re-ranking strategies and an advanced large language model (LLM) that reflects on query and evidence context, DeepSights thoroughly ensures the quality of and identifies the most relevant chunks of content before it's passed onto the user.

For the full, in-depth webinar discussing KM and AI, you can view an archived version of the webinar here.