What Autonomous AI Engineers Mean for Core Banking Modernization
Autonomous AI engineers are software agents that can independently plan, write, test, and deploy production code across large, complex systems to accelerate legacy system upgrades and core platform modernization. In banking, these AI agents enterprise deployments are emerging as a way to speed up core banking modernization while coping with aging infrastructure and scarce specialist talent. Instead of acting as simple code assistants, tools like Devin from Cognition behave more like full project contributors: they interact with existing repositories, run tests, and iterate on fixes without constant human direction. For institutions that depend on stable, highly regulated core platforms, this shift promises faster delivery of new features, security updates, and integrations, without rewriting systems from scratch. The result is a new model of development where AI agents handle detailed engineering work and human teams concentrate on product design, architecture, and risk decisions.
Inside Fiserv’s Deployment of Devin for Legacy System Upgrades
Fiserv has partnered with Cognition to deploy Devin, an autonomous AI software engineer, to speed core banking modernization and shorten development cycles. Devin operates directly inside banks’ complex codebases, where it can plan work, write and test code, iterate on bugs, and ship production changes using the same tools human developers rely on. According to Finovate, firms such as Goldman Sachs, Ramp, Zillow, and Lowe’s already use Devin to extend engineering capacity, suggesting the model is mature enough for critical infrastructure work. Fiserv will apply Devin to its core platform and other complex engineering initiatives so it can release new capabilities and integrations for thousands of client institutions more quickly. By treating AI agents as part of the engineering organization rather than a side experiment, Fiserv aims to make core upgrades and legacy system upgrades feel more like continuous improvement than occasional, high‑risk overhauls.
Faster Release Cycles and the New Developer Workflow
In traditional core banking modernization programs, release cycles stretch into months and large upgrades can consume years of effort. Autonomous AI engineers are changing that rhythm. Devin can maintain a constant pace of development work, from routine refactors to security patches, so human engineers can focus on architecture, design, and complex reviews. Fiserv Co‑President Dhivya Suryadevara said, “With Devin, we can accelerate modernization of the platforms our clients run their business on, ship new capabilities faster, and free our teams to focus on the work that matters most.” In practice, this means AI agents take on tasks like integrating new APIs, updating internal services, and improving quality checks, while human developers oversee requirements, compliance, and customer‑facing features. Over time, this blended workflow could make banks’ core platforms feel more adaptable, with smaller, more frequent releases instead of rare, disruptive upgrades.
Addressing Talent Shortages and Complexity in Banking Technology
The banking sector faces two linked problems: a shortage of skilled core systems developers and rising complexity in aging platforms. AI agents enterprise deployments promise relief on both fronts. Because autonomous AI engineers can work at scale across large, intertwined codebases, they are suited to the slow, resource‑intensive nature of legacy system upgrades. Russell Kaplan, Co‑Founder and President of Cognition, described Fiserv as the kind of organization where Devin creates “compounding value” due to its massive scale and ambitious engineering goals. Instead of hiring enough specialists to touch every subsystem, banks can use AI agents to extend their teams, reserving human attention for strategy, risk, and oversight. Fiserv’s decision to strengthen governance and security controls around AI‑assisted development underscores the point: success depends not only on faster code, but also on controlled, auditable processes that fit banking regulations and reliability expectations.