#31 The Future of Two-Pizza Teams with ChatGPT: A Call to Adaptability

#31 The Future of Two-Pizza Teams with ChatGPT: A Call to Adaptability

<< Previous Edition: Coding in the ChatGPT Era: The Rise of Code Reviewers

After attending DevOps Days Salt Lake City, I've been pondering how generative AI is going to reshape two pizza teams. As I've previously written, I believed team structures will consolidate over time, with the rise of roles such as 10X SRE and 10X DevOps engineer. However, recent conversations got me to get back on the drawing board and start thinking from scratch.

Eager Early Adopters vs Entitled Resistance

At the conference, I received mixed feedback from engineers. Some were excited and curious about the brave new world of generative AI, eager to adapt and not be left behind. On the other hand, there were those who were content with their current entitlement and unaware that their titles and functions may soon become obsolete.

Resistance and power plays may arise, but history has shown that they rarely succeed in halting progress. Just like how horse carriage owners in New York were unable to stop the rise of cars a century ago, those who resist the changes brought about by generative AI will likely face the same fate.

Two Pizza Teams

Let's dive into the topic of two pizza teams. Although they remain a goal for many organizations, most have not yet achieved them. Two pizza teams merge all roles into the team, including quality engineers, SREs, UI developers, and data engineers. With everything combined, it becomes easier to redistribute which roles will be handled by AI and which ones will remain with human beings.

I had previously predicted that the SRE role would merge with two pizza teams, but my perspective has shifted. While I believe some observability functions may still be handled by the team, the trend of creating separate SRE organizations will continue to accelerate.

Testing Redefined: Delegating to AI

As AI becomes increasingly capable, testing will likely be delegated to it with only occasional assistance from humans. At Roost.ai, we're focusing on helping developers automate the entire testing life-cycle, including unit-tests, service-2-service integration tests, and acceptance tests. Our goal is to empower developers to become 100x developers by streamlining the DevOps process with our chatGPT driven platform.

Data Engineers Take the Driver's Seat

The role of the data engineer will likely become the most powerful in the two-pizza team structure for several reasons. Firstly, the rise of machine learning models has made "walking on water" data scientists less relevant. Secondly, the roles of data scientists and data engineers have been merging into a 10x data engineer role. Finally, with the majority of test cases and a significant amount of code being written by GPT, the only remaining task is to deal with the data, making the role of data engineer critical for success.

Conclusion

As generative AI reshapes the landscape of software development, it will reward those who are curious and adaptable while punishing those who cling to outdated ways of working. It's up to each individual to decide which category they want to be in. With the rise of AI, the role of the data engineer is set to become more powerful than ever, while the structure of two pizza teams will continue to evolve to reflect the changing landscape. At Roost.ai, we're excited to be at the forefront of this revolution and are dedicated to helping developers thrive in this new era of software development.

>> Next Edition: Scaling Agile with the Power of ChatGPT: Two-Pizza

Dipanshu Mansingka

Principal Consultant / NITI's AIM/ATL Mentor

8 个月

can data engineer decide on filtering the dependent variables, which variable is categorical. what method to apply to find the dependent variables, I learnt both, practically worked more on explanatory BI (migrating etl from legacy to big data and report generation) then exploratory ai. Though i know all the methods for forecast, classification, optimization, svm,... i am not confident on which one to apply when.

回复

Hi Rishi, Bold predictions! ???? You say “[some] titles and functions may soon become obsolete.” Any timelines?

Rishi Yadav

Founder & CEO at Roost.ai

1 年

What I found most interesting after I finished my analysis was how data engineer will be the last person standing.

回复

要查看或添加评论,请登录

社区洞察

其他会员也浏览了