How to Drive Innovation & Democratize AI as a CS Leader ft. David Karp (Disqo)

In this episode, David Karp, CCO at Disqo, walks through the exact playbook for the Hackinator: a company-wide AI hackathon that turned 70 employees across a 300-person company into AI builders. Three to four of the prototypes are in production today. One is now being monetized with customers. And the CS team ran their own version without being asked.

---

Listen on YouTube, Spotify and Apple Podcasts

---

Three Things We’re Taking Away:

1. Remove the engineering constraint and participation explodes. Traditional hackathons are gated by who can code. Disqo’s Hackinator was open to anyone who could imagine a problem and articulate a solution using AI. CSMs, marketers, and data scientists led teams. 70 people out of 300 participated actively.

2. Seed the ideas, then get out of the way. Leadership put together a starter list of business problems across customer journey, data, and onboarding before the event. That list gave people a mental model for what was possible. Teams took it from there, self-organized, and produced things leadership hadn’t anticipated.

3. The momentum plan is as important as the hackathon itself. After the event: enterprise AI licenses for everyone, foundational training, bi-weekly AI office hours drawing 50-60 people, and a dedicated cross-functional team meeting monthly to drive AI workflow adoption. The hackathon was a spark. The infrastructure kept it burning.

🎯 The [Un]Churned Take: The Tactical Playbook for an AI Hackathon That Actually Sticks

David Karp, Chief Customer Officer at Disqo, ran a company-wide AI hackathon about a year ago at a 300-person organization. Seventy people participated. Fifteen teams built working prototypes in two days. Three to four of those prototypes are in production today. This conversation is the playbook.

How the Hackinator Was Structured

The first decision Disqo’s leadership made was that they had to go first. “We had to all agree it was important and that we were going to go first,” David said. That prior experience using AI themselves gave them credibility when they went to the rest of the company and said let’s go.

From there, the structure was deliberately simple. Leadership seeded a list of potential problem areas: customer onboarding, AE-to-CS handoffs, data trend discovery, sales-to-customer journey process flows. People could build on those ideas or submit their own. Teams self-formed, named themselves, and cross-functional pairing was encouraged but not mandated. The constraint that changed everything: you didn’t need to be an engineer to participate. If you could imagine a solution and work with AI, you had a seat at the table.

The Rules That Focused the Work

Three parameters kept the work grounded. Solutions had to create impact within six to twelve months. They had to use AI in some meaningful capacity. And teams had to produce a working prototype, not just a pitch. Scoring covered business impact, deployability, and creativity. Prizes were real: the winning team walked away with around $2,000, second place got $1,000, third got $500. The winning team also presented to the entire company at an all-hands.

Fifteen teams submitted valid ideas. Eight were close enough that judges went back and asked for clarification before scoring. “The scoring was so close that we actually went back to some teams and said, hey, can you clarify this piece.” The top three were all genuinely competitive.

What the CS Team Did Next

The winning team was CS-led. They had been hearing from customers about a specific capability Disqo didn’t prioritize on the product roadmap because it wasn’t prevalent enough. So they asked: what if we applied AI against our data sets to create some version of that solution without taking engineering time? They built it. Then hit a blocker. So they ran their own mini hackathon within the hackathon to solve the blocker. What they built is now being monetized with customers.

And then, a few weeks later, David found out the CS team had started running their own internal hackathon, unprompted. “I didn’t even know they were doing it until they had been doing it a couple weeks in and somebody told me.” That’s the sign a cultural shift actually took hold.

How They Kept the Momentum

After the event, Disqo did three things. They gave everyone enterprise access to a curated set of AI tools with training and data policies. They launched bi-weekly AI office hours where 50 to 60 people would show up to do show-and-tell, ask questions, and get paired with someone who could help. And they built a dedicated cross-functional team that meets monthly to drive AI workflow adoption across the business. The hackathon was a moment. The structure around it is what made it last.

Wrapping Up

You don’t have to coordinate a 70-person hackathon to get started. You have to get comfortable yourself first, find one colleague who’s dreaming about a problem, and start there.

See you next week 🧠

---

Where to find David Karp:

• LinkedIn: https://www.linkedin.com/in/davidalankarp/

---

Referenced in This Episode

• Disqo — The audience insights platform where David serves as CCO, focused on measuring advertising impact for brands. disqo.com

• Gamma — AI-powered presentation builder mentioned briefly as a tool teams used to put together hackathon pitches. gamma.app