Context Files: Scale GenAI Across Complex Work

Hosted by Kevin Dewalt

Fri, Mar 13, 2026

5:30 PM UTC (30 minutes)

Virtual (Zoom)

Free to join

Invite your network

Go deeper with a course

The GenAI Execution Framework: A Two-Week Intensive for Senior Executives
Kevin Dewalt
View syllabus

What you'll learn

Understand core shift behind context files

Learn why durable context improves AI work far beyond one-off chat prompts.

Build your first context file in 10 minutes

Create your own repeatable workflow with only Microsoft Word + an LLM

Design a scalable reasoning system

Prepare to extend this pattern to multiple projects and complex enterprise work.

Why this topic matters

Most GenAI stays shallow because chat has no durable memory. Teams chase tools, but the real bottleneck is how leaders structure problems and updates. Context files give GenAI stable, editable context. Using Microsoft Word and an LLM, you’ll build one in 10 minutes and leave with a repeatable workflow you can apply to any complex work.

You'll learn from

Kevin Dewalt

Three decades building enterprise AI — from neural nets to GenAI

I’m Kevin Dewalt, co-founder of Prolego. For three decades, I’ve built and led enterprise AI systems, from neural networks at Stanford in the early 1990s to today’s production-grade GenAI platforms. In dozens of Fortune 500 engagements, I’ve seen the same pattern: organizations do not lack ideas. They lack focus, sequencing, ownership, and adoption discipline.

My work centers on imposing that discipline. Inside enterprise AI CoEs, I help leaders narrow to the few initiatives that matter, align stakeholders, secure funding, and ship production systems. Over time, this became a repeatable approach for creating focus and execution under pressure.


Career highlights

  • Drove approval of $1–$10M annual enterprise AI budgets
  • Architected AI strategies underpinning $100M+ board-level investment decisions
  • Led dozens of AI strategy and engineering initiatives for Fortune 500 and highly regulated organizations, including Morgan Stanley, Citi, Bristol Myers Squibb, RTX, Bosch, FINRA, Lockheed Martin, Collins Aerospace, and others
  • Trained 1,000+ senior executives through private institutional AI leadership programs
  • Designed and facilitated executive AI workshops at Stanford, Carnegie Mellon, UC Berkeley, and MIT
  • Early strategic investor in Palantir at IQT; supported deployment of its first AI product in the U.S. intelligence community
  • 6x technology founder; investor in 35+ startups
  • M.S. Engineering (AI), Stanford; #1 graduate, U.S. Coast Guard Academy


Sign up to join this lesson

By continuing, you agree to Maven's Terms and Privacy Policy.