Checking in on Accrete: Building a Dual-Use AI Company That Scales
About Brian Drake & Bill Wall
Brian Drake serves as Federal Chief Technology Officer at Accrete AI, bringing deep government intelligence experience from his time at Defense Intelligence Agency (DIA). Bill Wall leads federal operations, managing the complex dance between classified and unclassified work across government contracts. Together, they've built a successful dual-use AI company that serves both commercial and government markets effectively.
About Accrete AI
Accrete AI develops decision analytics platforms that process massive information flows to identify patterns and predict outcomes. Their technology began in the commercial music industry, helping record labels identify viral talent from millions of social media posts. This same capability now serves government customers for information operations, supply chain risk assessment, and cognitive warfare applications. The company recently secured a $15M strategic financing agreement from the Air Force.
Key Takeaways
1. Matrix Organizations Enable Classified-Unclassified Collaboration: Accrete solved the clearance problem by building a "matrix organization" where cleared engineers on the federal side collaborate with corporate engineers without sharing classified information. "We had to build this dance between corporate and federal where cleared engineers get enough information to modify the product for classified networks without telling corporate engineers anything classified." This structure enables innovation while maintaining security boundaries.
2. Commercial Success Validates Government Solutions: Having a profitable commercial product dramatically improves government credibility. "One of the things that helped our IO product gain interest on the government side was the fact that it had been successful on the commercial side," Wall explains. Government buyers need proof that technology works in real-world conditions, and commercial contracts provide that validation better than demos or prototypes.
3. Sell Outcomes, Not Technology The biggest internal challenge was forcing technologists to focus on results rather than capabilities. "Nobody wants to buy AI. They want to buy outcomes." Instead of explaining how their neural networks work, they now lead with business impact: time savings, better decisions, reduced risk. "If you're explaining, you're losing" became their internal mantra.
4. Live Demos Beat Slide Decks Every Time: Brian’s experience sitting through 200+ vendor demos taught him that "about 70% of the market is full of s**t." The companies that stood out were those willing to run live demonstrations with audience-generated prompts. "Give me the name of a company. I'm going to run it through our system right now. They get to set the question, not me." This approach instantly separates real products from vaporware.
5. Bridge the Imagination Gap with Practical Questions: Government customers struggle to envision new capabilities they've never used before. The most powerful question Drake learned from a general officer: "What are you not doing that you should be doing?" This frames technology discussions around missed opportunities rather than current processes, helping customers see beyond incremental improvements to transformational possibilities.
For more on Accrete: www.accrete.ai
For more Crossing the Valley: valleycrossers.com
Share this post