Agentic AI that makes elite GTM teams superhuman | Frank Wittkampf

Ever wonder how massive sales forces at companies like Microsoft, Salesforce, or Databricks manage to consistently hit their targets and understand complex customer needs? A huge part of the answer lies in sophisticated, AI-driven insights, and today's guest is right at the heart of building that technology.

Today, I'm joined by Frank Wittkampf, Head of Applied AI at DataBook. DataBook is a platform designed to supercharge enterprise sales productivity. They don't just offer generic AI; they build deeply specialized systems that analyze vast amounts of data – financial reports, news, competitive landscapes, even proprietary insights – to tell salespeople exactly what to position, why, and when.   

They are moving beyond simple chatbots or free-form AI agents. DataBook focuses on applied AI, using what Frank calls 'guided reasoning' to ensure the insights delivered are consistent, reliable, and directly drive sales outcomes, like significantly increasing deal sizes.   

In this episode, Frank dives into how DataBook's AI works, why a 'guided' approach beats pure agentic systems in enterprise, the surprising challenge of people over-imagining AI's current capabilities, how they navigate the R&D frenzy to deliver real value, and their vision for a future where AI proactively coaches you.

Takeaways

• Why "Guided Reasoning" Beats Pure AI Agents: Enterprise needs predictable, repeatable outcomes - not creative exploration
• The Over-Imagination Problem: Why Computer Use and other flashy AI features aren't ready for enterprise deployment
• Data Strategy That Works: How Databook combines public data, proprietary datasets, and pre-solved analysis for instant insights
• R&D vs Reality Balance: Practical framework for experimenting with cutting-edge AI while delivering customer value
• The Future is Proactive: Why the next leap in AI isn't just responding to queries, but actively coaching users
• Enterprise Integration Challenges: Real talk about data access, security approvals, and building trust with large customers

Sound Bites

• "Free reasoning is all fun and great, but in enterprise, fully free reasoning is just not that helpful."
• "Computer use is incredibly useful... The problem with it is it's just so non-practical at the moment. It is incredibly slow."
• "For AI to deliver you a proper answer, you actually need to pre-solve that answer pretty thoroughly if you want to do a good job at it."
• "We can see deal sizes increase by 1.9 to 2x when people are actively using this."
• "The big change that's coming in AI is not just you engaging with it, but it engaging with you and helping you."

Chapters
00:00 - Introduction to Databook and Enterprise AI Reality
03:08 - What is Databook? Serving Microsoft, Salesforce & Databricks
04:33 - AI-Native Features: Beyond Simple LLM Implementations
06:17 - Customer Deep Dive: Why Big Tech Companies Choose Databook
09:18 - Proprietary Data Strategy and Pre-Solved Analysis
11:03 - Day-to-Day as Head of Applied AI: Product to Engineering Translation
14:21 - Balancing R&D Innovation with Customer Results
18:58 - Testing and Experimentation in Enterprise AI
21:14 - Dogfooding: How Databook Uses Its Own Product Internally
23:24 - What's Next: The Push Toward 4x Deal Size Increases
25:12 - Guided Reasoning: The Middle Ground Between Workflows and Agents
26:19 - Biggest Roadblocks: Enterprise Speed and Data Integration
27:49 - Technical Deep Dive: Delta Lake and Joint Data Access
30:07 - What Frank is Most Proud Of

Connect with us
Where to find Anthony:
LinkedIn: https://www.linkedin.com/in/wittkampf/
Medium: https://medium.com/@frankw_usa
Website: https://databook.com/

Where to find Sani:
LinkedIn: https://linkedin.com/in/sani-djaya/
Get in touch: [e-mail]