ABOUT FIBI

The First International Bank of Israel, FIBI, is one of Israel’s leading banking groups, providing retail, commercial and corporate banking, capital markets, investment advisory and private banking services. With a large in-house technology organization and a complex mix of legacy and modern systems, FIBI continuously invests in digital innovation to improve operational efficiency, accelerate development, and enhance customer services.

CHALLENGES

FIBI identified its technology and R&D organization as one of its largest cost centers and saw a clear opportunity to improve developer productivity, operational efficiency, and time-to-market.

At the same time, introducing AI into such a large and complex development environment presented several challenges:

• A multi-layered technology stack combining legacy architectures, modern platforms and system redundancies.
• A highly diverse development organization, bringing together Senior and Junior developers and highly experienced mainframe and COBOL engineers.
• The need to drive consistent AI adoption across teams with different levels of experience, technical environments and long-established development workflows.
• The risk of low adoption if AI tools were introduced through licenses and generic training alone.
• The need to demonstrate measurable business value while building sustainable, organization-wide adoption.

Following a market evaluation, FIBI selected Commit AI for its practical approach, deep understanding of enterprise development environments and ability to tailor End to End AI R&D Transformation journey and Agentic SDLC to the bank’s specific needs.

SOLUTION

FIBI selected Commit AI to lead a hands-on AI transformation across its R&D organization.

At the core of the initiative was the implementation of GitHub Copilot, positioned as part of a broader move toward an agentic Software Development Life Cycle, where developers increasingly work alongside AI throughout the development process, augment and delegate tasks seamlessly.

A key component of the program was Commit AI’s Forward Deployment Engineer model. Dedicated AI transformation experts were embedded full-time within FIBI’s technology organization, working directly alongside development teams. Their day-to-day presence gave developers immediate access to expert guidance within real projects, helping build confidence, accelerate adoption and turn AI into a natural part of everyday development work.

Their role included:

• Delivering customized workshops based on real development scenarios.
• Providing one-on-one mentoring and hands-on support.
• Joining active bank projects and embedding AI into everyday workflows.
• Building tailored AI skills and agents for specific teams and use cases.
• Supporting both legacy and modern development environments.

The rollout was phased to ensure adoption before scaling:

Phase 1

50 developers supported by 2 Deployment Engineers.

Phase 2

An additional 250 developers supported by 3 additional Deployment Engineers.

The program ultimately reached 300 developers supported by 5 full-time Commit AI Deployment Engineers.

To ensure visibility and accountability, Commit AI and FIBI also established monthly steering committees and performance dashboards to track adoption, active usage and productivity improvements.

RESULTS

Partial examples

• 300 developers onboarded: The initiative scaled from an initial 50-person pilot to 300 developers across the organization.
• 18 months reduced to 8 months: A complex modernization initiative consolidating four projects saw its projected development timeline reduced by approximately 10 months.
• Approximately 50% faster development: A legal reporting system was delivered in roughly half the original development time.
• Tens of thousands of development hours saved: Across multiple projects, the initiative generated significant productivity gains and savings worth millions of shekels.
• Broad organizational adoption: AI adoption gained strong support across the development organization, including among highly experienced COBOL and mainframe engineers, who became active users and internal champions.
• Unified developer experience: FIBI established a more consistent AI-assisted development environment across both legacy and modern technologies.
• Stronger AI adoption across R&D: Rather than remaining an isolated tool rollout, AI became embedded into everyday development processes and active projects.
• A scalable AI operating model: FIBI established a repeatable framework for expanding AI-assisted development across teams, technologies and future use cases.
• A strategic partnership: Commit AI evolved from a technology provider into a trusted partner working closely with FIBI on its broader R&D transformation.
  ‍

“Commit AI stood out because they recognized that enterprise AI adoption requires more than generic online courses. Through their full-time Deployment Engineer model, they embedded expert mentors directly onto our floor, combining hands-on guidance with governance dashboards to unify our developer workbench across our entire legacy stack and hit our targets.”

Alon Sandak, CTO, FIBI

‍