DiscoverBeyond The Pilot: Enterprise AI in ActionHow JPMorgan Engineered a 30K AI Agent Economy
How JPMorgan Engineered a 30K AI Agent Economy

How JPMorgan Engineered a 30K AI Agent Economy

Update: 2025-12-17
Share

Description

Inside the 'Agent Economy': How 30,000 AI Assistants Took Over JPMorgan


While most enterprises were scrambling after ChatGPT launched, JPMorgan Chase was already two years ahead. 🚀


In this episode of Beyond the Pilot, we sit down with Derek Waldron, Chief Analytics Officer at JPMorgan Chase, to reveal how the world’s largest bank built an internal AI platform that is now used by 1 in 2 employees daily.


Derek shares the contrarian insight that drove their strategy: AI models are commodities; the real moat is connectivity.


Learn how they scaled from zero to 250,000+ users, why they empowered employees to build 30,000+ of their own "Personal Agents," and how they are solving the data privacy challenge at an enterprise scale.


🔥 IN THIS EPISODE:




  • The "Super Intelligence" Thought Experiment: Why raw intelligence is useless without enterprise connectivity.




  • The Agent Economy: How JPM enabled non-technical staff to build 30,000 custom AI assistants.




  • The Adoption Playbook: How to break through the "30% wall" and get the majority of your workforce using AI.




  • Build vs. Buy: Why JPM built their own "LLM Suite" instead of waiting for vendors.




⏳ CHAPTERS:


00:00 - Introduction: The JPMorgan AI Story


01:45 - The 3 Core Principles Behind JPM’s Strategy


03:25 - The "Super Intelligence" Thought Experiment


05:00 - Data Privacy: Why JPM Doesn't Train Public Models


06:00 - Viral Adoption: From 0 to 250k Users


09:20 - Evolution of LLM Suite: From RAG to Ecosystem


14:00 - The "Moat" is Connectivity, Not the Model


23:00 - The Agent Economy: 30,000 Employee-Built Assistants


31:00 - Governance & Guardrails for AI Agents


33:00 - Crossing the Chasm: Getting to 60% Adoption


40:00 - The "Product" Mindset: Solving Business Problems First


42:30 - The Future: End-to-End Process Transformation


46:25 - The "Unsolved" Problem Derek Wants to Fix


🙏 SPECIAL THANKS TO OUR SPONSOR:


This episode is presented by Outshift by Cisco.


Learn more about their work on the Internet of Agents and the open-source Linux Foundation project:


đź”— https://www.agentcy.org


🎙️ GUEST: 


Derek Waldron | Chief Analytics Officer, JPMorgan Chase


HOSTS:


Matt Marshall | VentureBeat


Sam Witteveen | VentureBeat


#EnterpriseAI #JPMorgan #GenerativeAI #AgenticAI #FinTech #ArtificialIntelligence #Innovation #BeyondThePilot


.


.


.


Subscribe to VentureBeat: 


   /  @VentureBeat  


.


.


Subscribe to the full podcast here:


Apple: https://podcasts.apple.com/us/podcast/venturebeat/id1839285239


Spotify: https://open.spotify.com/show/4Zti73yb4hmiTNa7pEYls4


YouTube: https://www.youtube.com/VentureBeat


https://www.youtube.com/playlist?list=PLMQoSwszBxm5dCv2bdqGnJ0QAL9n7Ds4_


đź”— CONNECT WITH VENTUREBEAT:




Learn more about your ad choices. Visit megaphone.fm/adchoices

Comments 
loading
In Channel
00:00
00:00
1.0x

0.5x

0.8x

1.0x

1.25x

1.5x

2.0x

3.0x

Sleep Timer

Off

End of Episode

5 Minutes

10 Minutes

15 Minutes

30 Minutes

45 Minutes

60 Minutes

120 Minutes

How JPMorgan Engineered a 30K AI Agent Economy

How JPMorgan Engineered a 30K AI Agent Economy

VentureBeat