AI for Engineering Leaders: How I Finally Focus on Real Leadership

The tech world buzzed this year with endless discussions about AI transforming software development. Every day, new headlines appeared about developers boosting productivity with GitHub Copilot or debating whether or not AI would replace developers. Yet, in this flood of coverage, I did not see much about how AI can revolutionize engineering leadership and transform how we manage technical teams.

Before we dive in, take a moment to reflect:

• How many hours do you spend each week just gathering context for meetings?
• How often do you wish you had better notes from past 1:1s?
• How do you stay on top of industry trends across multiple technology domains?
• When was the last time you felt truly prepared for every meeting in your day?

I have spent the past year encouraging my teams to include AI tools like GitHub Copilot in their daily routines, as I deeply agree with the GitHub blog post saying AI won’t replace developers, but developers using AI will replace those who don’t. Then it struck me: if I believe this will happen to developers, I should really consider this happening to tech leaders too! AI’s potential to transform engineering leadership might be even greater than its impact on coding.

The Hidden Challenge of Engineering Leadership

Engineering leadership is unique in its breadth. On any given day, I might jump from reviewing architectural decisions to mentoring team leads, from analyzing performance metrics to planning strategic initiatives, from exploring promising new tech to helping the team solve an incident in production. This broad scope creates a significant challenge: the sheer amount of preparation and context-gathering required to make informed decisions.

How much of your week is spent on actual leadership versus preparing to lead? Think about it. Before that architecture review meeting, you need to understand the latest discussions across multiple GitHub issues and Slack threads. Before your 1:1s, you should review past conversations and team member contributions. To guide your team effectively, you must stay current with industry trends, security threats, emerging technologies and understand the current bottlenecks of your teams. And to grow your organization, you need to track speaking opportunities, relevant conferences, and potential learning resources for your team.

This preparation consumes hours of a leader’s week. But I believe none of this preparatory work is where we provide our real value. Our true impact comes from what we do with this information: the insights we draw, the connections we make, and the guidance we provide.

This is where AI enters the picture.

Transforming Leadership with AI: My tools and workflows

Industry Research and Discovery with Perplexity AI

Perplexity AI is an AI-powered search engine that adds to traditional search by synthesizing information from multiple sources and presenting coherent, contextual answers. Unlike traditional search engines that return a list of links, this AI tool provides direct answers while citing its sources, making research much more efficient and allowing you to access the underlying articles if you want to.

How long did you spend last week browsing tech news? What if you could get a curated summary in minutes? This is what I use Perplexity AI for. Here are a few examples of what I ask to Perplexity AI frequently:

• Track industry trends and emerging technologies;
• Find relevant conferences and speaking opportunities;
• Discover training resources for my team;
• Monitor our competition and market dynamics.

With a single prompt, I can get a comprehensive list of all conferences with ongoing calls for papers related to WordPress, PHP, Node.js, Django, or Web performance in Europe. Instead of manually browsing through various event websites and checking each conference’s status, I get an instant, curated list of opportunities for my team and myself.

When was the last time you searched for speaking opportunities? How many tabs did you have open trying to track submission deadlines?

I also have a weekly ritual of asking Perplexity for recent articles related to the WordPress ecosystem and trends in our technical stack. Rather than visiting dozens of blogs and websites, I get a curated list of the most relevant content. I can identify quickly what’s worth a deeper read, saving me hours of manual browsing while ensuring I don’t miss important developments in our industry.

Internal Knowledge Processing with Dust.tt’s assistants

Dust.tt is an AI-powered platform that helps teams make sense of external and internal knowledge. What makes it particularly powerful is its use of RAG (Retrieval-Augmented Generation) technology. RAG combines the power of large language models with the ability to retrieve and reference specific information from your organization’s data. This means it doesn’t just generate generic responses – it provides insights specifically grounded in your team’s context and history.

Think about your last performance review preparation. How many hours did you spend reviewing old notes and messages?

Think of it as having a personal assistant who reads and summarizes all your important documents and conversations, and who can also connect dots across months of interactions while maintaining perfect accuracy about what was said or documented. By connecting to various data sources (like Microsoft Teams, Notion, Slack, GitHub, documents, and other workplace tools), Dust.tt creates a searchable, AI-powered knowledge base of your organization’s information.

How confident are you that you remember all the key points from last month’s 1:1s?

Managing multiple teams means processing vast amounts of internal information. I use Dust.tt to:

• Generate summaries from Microsoft Teams 1:1 transcripts,
• Analyze sprint retrospectives for patterns,
• Process team feedback and identify improvement areas,
• Generate activity summaries from release notes, GitHub activities, and Slack discussions,
• Maintain context across multiple projects.

The game-changer here is the automated parsing of internal data and notes. My 1:1s are automatically transcribed, summarized, and stored. The same goes for sprint retrospectives. When I need to assist the team in prioritizing technical tasks for cooldown sprints, we can refer to a curated summary of priorities from those notes. When I need to review someone’s progress or prepare for a performance review, I have AI-generated summaries of our entire conversation history at my fingertips.

Because of RAG, these summaries are always grounded in actual conversations rather than AI hallucinations, making them reliable for important decisions. The assistants help me retrieve the specific information I need at a given point in time so that I can easily remember it and then make informed decisions. This automated monitoring helps me to efficiently refresh my memory before I walk into a meeting without spending hours manually gathering information.

One key feature of Dust.tt is that it comes with an assistant to write effective prompts and help you build your own assistants! I simply discuss with this prompt engineering assistant about what I would need, and it helps me to build a custom assistant for it. The results are very impressive! This taught me a lot about prompt engineering, but it is also a great introduction to the principle of agentic AI and chaining AI assistants together. Playing with those new tools is a great way to stay on top of the trend!

Real Impact on Leadership Effectiveness

The impact of productivity has been quite impressive for me and led to impacts going beyond just saving time.

The most significant change has been in my strategic focus. Instead of drowning in preparation work, I can now fully dedicate my time to taking action, getting things done, strategic thinking, and creative problem-solving. With AI handling the heavy lifting of information gathering and processing, I’ve found myself better able to spot patterns that I might have missed or forgotten before or to engage in technical discussions and support my teams more effectively. I believe the saved time overall leads to better strategic decisions.

The quality of my meetings has transformed as well. I now walk into conversations more prepared, with a refreshed memory and a stronger context awareness, which allows me to directly dive into much more meaningful discussions. The follow-up process has also improved dramatically, thanks to AI-processed meeting summaries that capture not just what was said but also help to highlight the underlying themes and action items.

Overall and most importantly, I’m more present in my interactions. In the past, I often found myself splitting my attention between actively listening and taking comprehensive notes during 1:1s. Now, I can focus entirely on the conversation and the person in front of me because I am already up-to-speed and I know that AI will help me process and retain the important points. This has led to deeper, more meaningful, and productive conversations with my team members.

My Journey with AI Tools

When I started exploring AI tools for my role, I began small. I identified my biggest time sink – staying updated with industry trends and finding opportunities for my team – and started experimenting with Perplexity AI. The results were encouraging enough that I gradually expanded to other areas of my work.

Trust didn’t come immediately. I spent time verifying AI-generated summaries against my own notes, and I still do for critical information. But over time, I’ve learned where AI excels and where human judgment remains essential. For instance, I trust AI to summarize a technical discussion, but I always rely on my own insight for personnel decisions and strategic planning.

The tools I use today, like Perplexity AI and Dust.tt, work well for my needs. I continue to experiment with new tools as they emerge, partly because, as an engineer, I find this quite enjoyable. But also because I believe some game-changers are yet to appear. The AI landscape reminds me of the early days of smartphones – remember how many app stores and platforms existed before the field consolidated with go-to apps and leaders in each domain? We’re in a similar phase with AI tools, and staying curious has helped me find solutions that truly enhance my workflow.

This reminds me of earlier technological transitions. Just as those who mastered spreadsheets early became more productive, I believe those who learn to effectively use AI tools now will have a similar edge. I don’t believe it’s about automating leadership – that’s impossible and missing the point. It’s about empowering us to focus on what truly matters: leading and developing our teams.

The future of engineering leadership isn’t about replacing human judgment with AI. It’s about finding the right balance between AI-powered efficiency and human insight. While we’ve been focusing on how AI will change coding, it’s already transforming what it means to be an effective engineering leader. The question isn’t whether to embrace AI in your leadership practice, but how to make it work for your unique situation and team.