Adopting Copilot without adjusting your Agile methods is a mistake

Agile organizations have outperformed others in making and managing decisions faster. Over the past two decades, Agile adoption in IT has introduced changes in the way infrastructure, applications, data, and skills are produced and/or consumed. The foundational elements of agile, such as collaboration, automation and continuous improvements have been THE innovation sources for newer methods and tools for application development and deployment.

Two recent events have disrupted the agile progression in IT. One is the Covid-led? era of flexibility in how, when and where we work, challenging the ways we collaborate.? The other one is the promising Generative AI and the associated complexities to the IT governance. The incremental economy that these events have created forces every company to take the opportunities and the challenges seriously.

Distributed Agile is in practice for quite some time. The remote/hybrid work realities are just extension to what we have already seen in distributed teams, whereas the productivity promises of tools such as Copilot X etc. are new. Assuming that the current Agile practices would just work with GenAI practices is a mistake. So, what can we do?

Here are some of my thoughts. Add yours.

1. Extend DevOps to include representation from DataOps and MLOps. Given the importance of data and AI/ML models, and their own nurtured lifecycles in the businesses, it is critical that DevOps teams should have representation from the DataOps and MLOps teams as well (ModelOps is a subset of MLOps). Only then the objective of bringing the "production" and "operations" closer can be achieved.
2. Software intelligence is more important than ever. Not understanding the application systems holistically, before auto-producing the code in production will be disastrous. Corporate IT is a string of AI and Non-AI mix of applications and IT assets. Additionally, "explainability" of the code can be achieved, only when we have the software intelligence on the code that GenAI produced. End of the day, it is not the functional correctness, but the architectural fitment that matters most for unlocking productivity improvements.
3. Continuous Compliance and Continuous Security are equally important. One of the main concerns of GenAI tools is the vulnerabilities that auto-generated code can introduce to the corporate IT. Now, it is important to make proper adjustments at the pod-level for compliance and security, so that they are designed and delivered continuously, rather than audited and assured periodically.
4. Augmenting the quality gates in your CI/CD pipe for AI assistants. The founding principles of opensource - transparency, inspection, and adaptation - can be extended to GenAI products as well. "Inspection" should not only cover the quality, performance, security and UX aspects of the code that the tools provided, but also the architectural fitment in the enterprise IT.
5. Skills mismatch will impact productivity promises. Note that the tools are as good as its handlers. An experienced developer can demonstrate the above-average productivity level with an AI assistant, but an inexperienced can create more problems than benefits. Training the dev/QA community to handle the tools and the guidelines for the code/test governance will require adjustments to the Agile operating model. Also remember that DevOps team is more than developers, DataOps team is more than data engineers, and ModelOps team is more than data scientists. The cross-disciplinary skills mix of the devops teams will significantly change when AI is part of the conversation.

Adjusting the methodology for the perceived challenges should not curtail the potential benefits that GenAI can produce. If used properly, Gen AI can help in hyper-automation of dev/QA tasks, evaluating design options through rapid prototyping, simplifying the documentation process, monitoring production environment to predict performance bottlenecks, etc.

Finally, if we don’t adjust our agile methods to the above realities to unlock value at a faster pace, "time to market" and the associated "cost benefits" will be perceived poorly. Changes to agile methodology are inevitable because GenAI and Agile are providing real competitive advantages. Don’t take it easy.