Mapping the new SDLC at BNY: Codifying AI into every step of the delivery lifecycle (Jason Valentino)

Engineering Enablement by DX

0:00

-33:24

Mapping the new SDLC at BNY: Codifying AI into every step of the delivery lifecycle (Jason Valentino)

BNY’s AI strategy focuses on optimizing the entire engineering workflow, from planning to production.

Justin Reock

Jun 08, 2026

Listen and watch now on YouTube, Apple, and Spotify.

Jason Valentino is Head of Software Engineering Strategy at BNY, where he oversees developer tooling, DevEx, platform workflows, and software delivery governance across more than 8,000 engineers.

In this session from DX Annual, Jason shares how BNY moved beyond AI coding assistants to rethink the entire software delivery lifecycle. He explains how his team identified bottlenecks across the SDLC, prioritized automation opportunities, and applied AI to planning, peer review, testing, change management, and compliance workflows.

Jason also discusses what it takes to scale AI inside a highly regulated enterprise, including rewriting policies, partnering closely with risk and audit teams, and building a culture that encourages experimentation and rapid sharing of ideas.

Some takeaways:

Start with the 3X stress test

Ask what breaks if engineering throughput triples. Jason’s team began by assuming AI would dramatically increase the volume of pull requests, code reviews, and releases, then identified which systems and processes would become bottlenecks.
Map every step of the SDLC. BNY listed each task across planning, coding, testing, peer review, and release governance to understand which steps were manual, partially automated, or already well-instrumented.
Use developer sentiment to prioritize investments. By combining workflow analysis with DX survey data, the team focused on the areas causing the most friction rather than chasing the latest AI use case.

Apply AI in three distinct ways

Use IDE and CLI tools to amplify individual developers. Tools like Claude Code, Windsurf, and Codex help engineers move faster while still working within established guardrails.
Deploy autonomous agents for repetitive work. BNY’s “digital workers” handle tasks like access requests, backlog grooming, and other low-value activities that engineers would rather avoid.
Embed AI directly into workflows. The biggest gains come when AI is triggered automatically as part of code review, change management, and testing rather than relying on developers to invoke tools manually.

Use small automations to compound over time

Automate the tedious parts of planning. BNY added AI capabilities to Jira to draft stories and epics, lint requirements, and assign confidence scores.
Turn one automation into the next. Once a high-quality story exists, it becomes the foundation for generating test cases and other downstream artifacts.
Look for highly manual actions. Jason recommends watching how teams actually work and identifying repetitive tasks that are prime candidates for automation.

Rebuild governance for an AI-assisted world

Rewrite policies and controls. Existing language around code review, approvals, and software delivery often assumes humans perform every step and must be updated to reflect AI-assisted workflows.
Bring risk and audit teams in early. Rather than presenting finished solutions for approval, BNY collaborates with governance partners while designing new approaches.
Codify deterministic rules. AI can handle routine work automatically, while larger or riskier changes are routed to humans for additional oversight.

Treat duplication as a feature, not a bug

Expect multiple teams to solve the same problem. In a large organization, some overlap is inevitable when thousands of people are experimenting with AI.
Use show-and-tell to surface innovation. BNY hosts weekly sessions where teams demonstrate what they’ve built and share lessons learned.
Consolidate the best ideas. Once similar solutions emerge, platform leaders can combine the strongest features into shared capabilities.

Create a culture that rewards experimentation

Start saying yes. Jason’s advice to engineering leaders is to lower barriers and put promising ideas in front of users quickly.
Treat internal tools like products. Successful experiments are documented, shared, and iterated on rather than left as one-off hacks.
Make engineering fun again. For Jason, one of the biggest wins of the past year has been seeing teams energized by the opportunity to solve meaningful problems with AI.