Azure DevOps‑Based Agile and DevOps Transformation at Beijing Bank: Practices and Lessons
This article details Beijing Bank's 2017‑2018 journey of adopting Azure DevOps Server to enable a regulated yet agile development process, covering challenges, project‑, version‑, and test‑management solutions, gene‑adaptation of TFS, and the resulting dual‑mode, automated workflow for a large financial institution.
On October 27, 2018, at Microsoft’s largest annual technical conference in Shanghai, Zhou Bing, head of Beijing Bank’s channel system and agile team, presented the bank’s DevOps transformation experience, which received strong audience feedback.
Beijing Bank, founded in 1996 with over 620 branches, has been pursuing digital transformation; in June 2017 it introduced Azure DevOps Server (formerly TFS) as the core tool for agile rollout, aiming to improve responsiveness while complying with strict regulatory and process requirements.
The bank faced typical financial‑industry challenges: heavy regulation, multiple heterogeneous systems (channel, accounting, payment, data, bus), a high volume of projects and releases, complex environment and version management, and lengthy, document‑heavy workflows that created information silos.
To address these issues, the bank sought a platform that could manage the full lifecycle of projects, provide automated CI/CD pipelines, enable flexible version control, and unify development processes across project and system dimensions.
Azure DevOps Server was selected because it integrates project management, version control, continuous integration, continuous delivery, and test management in a single solution. The bank leveraged both the project‑level and system‑level capabilities of the platform.
During pilot phases with different system teams, the bank identified two major “collisions”: the idealized, agile‑centric design of TFS versus the bank’s regulated, process‑heavy reality, and the need for a simple, flexible tool that still fit existing workflows. Customizations and gene‑adaptation were applied to reconcile these differences.
For project management, the bank created a “project management” team project in TFS that automatically syncs requirements from the internal system and established a system‑level team project for each application, linking them via requirements to achieve clear cross‑dimensional governance.
Version management adopted a Git branching model with three branches—development, controlled, and production—corresponding to development, pre‑production, and production pipelines, thereby aligning version control with the bank’s release cadence.
Test management continued using the existing test‑management system but integrated defect flow into TFS, allowing developers to focus on requirement implementation and defect resolution within a single interface.
The practice summary highlights three distinctive features: a dual‑mode approach that balances stable (regulatory) and agile (innovation) streams; embracing internet‑style Git branching while maintaining rigorous risk controls; and extensive process automation (code security scanning, approval workflows, document checks, environment tracking) that boosts efficiency and quality.
In conclusion, Beijing Bank built an integrated toolchain that connects the entire development lifecycle, achieves fine‑grained management, centralizes data, enables flexible versioning, and automates quality checks, thereby enhancing unified collaboration and preparing for continued agile and DevOps advancement.
DevOps
Share premium content and events on trends, applications, and practices in development efficiency, AI and related technologies. The IDCF International DevOps Coach Federation trains end‑to‑end development‑efficiency talent, linking high‑performance organizations and individuals to achieve excellence.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.