How to Unlock the Knowledge Hiding Inside Your Firm
📰 Discover how architecture firm Shepley Bulfinch leverages disciplined experimentation and video-first documentation to enhance knowledge sharing. Their innovative approach to AI and culture empowers teams, making knowledge management effective without a dedicated department.
🎉🍾 Tomorrow is my birthday, which also marks the one year anniversary of this newsletter. While publishing 45 of these and writing over 41,000 words, I've discovered something unexpected: writing these deep-dive analyses has fundamentally changed how I listen to my own podcast. What started as a way to extend the conversation has become a tool for clarity. It’s forced me to identify the signal within the noise, to extract the frameworks hiding in casual dialogue, and to connect insights across episodes that I might have missed in real-time. The newsletter has become the complement I didn't know the podcast needed, transforming ephemeral conversations into permanent, actionable knowledge. I hope you are getting the intended value as well!
On to this week’s edition and my opening question for you:
What hidden value is sitting inside your firm right now buried in old videos, conversations, and workflows just waiting for AI to surface it?
Summary
In episode 211 I sat down with Jim Martin and Jess Purcell from Shepley Bulfinch, joined by our KM 3.0 guide, Chris Parsons, to get an inside look at what it really takes to make AI useful inside an architecture practice. Shepley is a 150-year-old firm with an experimental culture, and their approach to AI, video-first documentation, and knowledge-sharing offers a template for any AEC organization trying to modernize without losing its soul.
What emerged was a candid exploration of disciplined experimentation, the hard edges of AI’s unpredictability, the cultural scaffolding needed for knowledge to flow, and the reality of running knowledge management without a KM department, all while maintaining an environment that encourages curiosity rather than compliance.
Key Takeaways
Here are my top takeaways from the podcast episode. Then we'll get into the deeper analysis.
- Disciplined experimentation: Shepley defines value by continuously testing, retesting, and documenting their tools, not by believing the hype.
- “Trust but verify” is the new mantra: Teaching staff how to question AI outputs is now just as important as teaching them tools.
