1 hour ago
22
BNY Mellon gathered nearly 400 employees to watch pitches at an internal demo day, a sign that the 240-year-old institution is pushing AI adoption well beyond the pilot stage. The event reflects a broader initiative at BNY to embed agentic AI, meaning AI systems that can independently take actions and make decisions, into the daily workflows of one of the world’s largest custodian banks.
For a company that holds roughly $50 trillion in assets under custody, the move from experimenting with chatbots to deploying autonomous agents across trading, compliance, and client services represents a meaningful strategic shift.
Inside BNY’s AI agent factory
The backbone of BNY’s agentic push is Eliza, a proprietary AI platform the bank has built to let employees create, test, and deploy AI agents without needing a computer science degree. BNY has run bootcamp-style training programs through the platform, graduating thousands of employees who can now build their own agents. The bank has deployed somewhere between 100 and 200 AI solutions internally, spanning functions from asset management to regulatory compliance.
The platform got a significant upgrade when BNY integrated Google Cloud’s Gemini Enterprise into Eliza, boosting its capabilities in research automation and data synthesis.
At a recent NY Tech Week event, BNY showcased some of the practical applications coming out of this ecosystem. One standout was Agent Maven, an agentic automation tool designed for two specific use cases: proactively monitoring trade settlements and qualifying sales leads. The first catches potential settlement failures before they become expensive problems. The second helps relationship managers focus on the prospects most likely to convert, rather than burning hours on cold outreach.
The bottom-up innovation bet
What makes BNY’s approach distinct from the typical enterprise AI rollout is its emphasis on bottom-up enablement. Rather than having a centralized AI team hand down solutions from on high, BNY is training frontline employees to identify problems in their own workflows and build agents to solve them.
The demo day format reinforces this philosophy. Employees pitch their AI agent concepts to colleagues, creating an internal marketplace of ideas where the best solutions can surface organically. BNY maintains governance and compliance frameworks around the Eliza platform to ensure that creativity doesn’t outpace controls.
What this means for the financial sector
The integration of Google Cloud’s Gemini into BNY’s stack signals that major cloud providers are actively competing to become the AI infrastructure layer for traditional finance. BNY’s approach highlights an emerging reality: AI strategy is no longer a technology question, it’s a workforce question. The banks that figure out how to turn their existing employees into AI builders, rather than replacing them or hiring entirely new teams, will likely move faster and with fewer integration headaches than competitors pursuing top-down deployments.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
English (US) ·