My memo to the team about AI trend

The rise of AI is no longer just a hype, it’s reshaping the way we work, especially in engineering. Over the past year, I’ve heard a lot of conversations about whether AI will replace software engineers. I’ve said this before, and I still stand by it: AI won’t replace engineers. But it will change the way we work.

Being an engineer isn’t just about writing code. It’s about understanding problems, designing elegant solutions, and building things that matter. AI is a tool, not a replacement, and we can be much more effective if we know how to use it well. If we continue working the same way we did before AI, we can’t expect to move faster or build better.

In early April, I shared this memo with my team, and it is a reminder and a call to action: it’s time to rethink how we work, how we use AI, and how we level up together.

2025: Shaping the Future

Why does AI matter now?

The AI era is already here, and it’s changing how we build, test, and deliver products. Just earlier this month, Shopify’s CEO announced that using AI is now a baseline expectation for every employee, from executives to engineers. Before hiring anyone new, teams must prove that AI can’t do the job better. This bold move highlights a major shift happening across the tech industry: Companies that move fast with AI will lead, and those that don’t will fall behind.

We need to shift our mindset: AI isn’t just a tool; it’s our new teammate that helps us move faster and smarter.

What does it mean for us?

Leveraging AI

Using AI to handle repetitive tasks or refine your writing is the minimum expectation. But let’s not stop there. Our real goal is to deeply integrate AI into how we think and create. Building solutions with AI as our partner will define our future success. Whether it’s automating QA validation steps, generating test cases from API specs, creating dashboards from raw logs, or summarizing meetings and reports instantly. We should look at every problem space and ask: How can AI be an active builder with us?

Accelerating Our Delivery

In the next 6 months, integrating AI into your daily workflow is not optional; it’s expected. AI should reduce the time spent on understanding the problem & codebase, coding, documenting, and testing by 30-50%. This reduction should be reflected in project timelines and tracked through metrics such as actual lead time and test coverage, etc. If your project estimates remain the same, ask yourself and your team: Why aren’t we leveraging AI effectively?

Automation as the Norm

We’re aiming for full automation of manual workflows within the next 6 months. Managers and engineers need to proactively identify tasks that AI can handle or accelerate. For example, we already generate incident reports from chat messages in the incident channel. What else can we automate? Share your ideas with your manager.

The expectations

• AI is part of daily work. Every engineer is expected to use AI to assist with coding, testing, documentation, and communication.
• Speed up delivery. With AI, we should see a 30-50% reduction in the time spent on understanding the problem & codebase, coding, documenting, and testing. Project estimations should reflect this acceleration.
• Automate manual workflows. Think beyond what you’re doing today, identify steps in your workflow that can be automated using AI. Propose them to your manager.
• No more repetitive work. If a task is being done manually and repeatedly, it’s a signal for automation. Beyond incident reports, think about recurring release note updates, syncing meeting notes to task boards, or even cleaning up feature flags, etc. These are all opportunities for AI to take over; use AI to handle these tasks.
• AI-first mindset. When starting any new task or project, think: “How can AI help me do this better or faster?”
• Raise the bar. The new baseline includes using AI tools like Copilot, Cursor, Claude, ChatGPT, and others. If you’re not using them yet, start now. Learning and experimenting with AI is part of the job.
• Proactive contribution. Everyone should bring forward suggestions on how AI can improve the way we work. Your ideas matter.
• Hiring is the last resort. We only consider hiring new headcount when it is for a new skill or when it is truly necessary. Teams must first demonstrate why the work cannot be done effectively using AI. This should become part of our default thinking: build with AI first.

Let’s lead the AI era by example, embrace AI, rethink our workflows, and set a new standard for speed and innovation, just as our vision states: to be a world-class engineering team, proud of our craft, passionate about problem-solving, and always improving together.