Across performance reviews, design decisions, and strategic communication, AI tools can reduce overhead and amplify the impact of strong management. 

Ask any engineering manager what their day looks like, and you’ll hear variations of the same story: a packed calendar, dozens of Slack threads, a doc that needs feedback, a teammate who needs coaching, a hiring panel to prepare for. And somewhere in between, you’re expected to stay aligned with product, infrastructure, and leadership priorities, while making sure the team is executing reliably.

But the job hasn’t just stayed busy. It has grown in both scope and intensity.

As developers ship faster with the help of AI tools, the pace of change across engineering has accelerated. Code is written more quickly, features move into production faster, and iteration cycles have shortened. That shift means more planning, more design reviews, more cross-team alignment, and more people to support, all while maintaining a high quality bar.

Managers are now expected to lead larger teams, support faster-moving projects, and uphold high standards across both technical and organizational work. And this isn’t temporary. Big Tech companies have been intentionally reducing middle management layers. Google, Microsoft, Amazon, Meta, and others are flattening their structures. Google even cut around 10% of its manager roles as part of an efficiency push. 

That shift is only gaining momentum. In this new model, engineering managers are suddenly accountable for more direct reports, more coordination, and more decision-making, with outdated tools and less time.

In this landscape, making every manager more productive is no longer optional. It has become a business imperative that directly affects team performance and organizational resilience.

People and culture

People leadership is central to engineering management. Tasks like providing feedback, coaching, hiring, and supporting team growth demand sustained attention and emotional energy. While these responsibilities are deeply human, many of the workflows behind them are repetitive and time-consuming.

AI tools have not replaced judgment or decision-making. Instead, they have made these high-effort tasks faster, more structured, and easier to scale. The following examples reflect how I have applied these tools in my own day-to-day work, and how they might be helpful to others facing similar challenges.

Performance reviews and promotions

Performance reviews are about continuously understanding how engineers are driving impact across code, design, collaboration, and support. As teams grow and engineering productivity increases, managers need better tools to answer nuanced questions like:

“Who consistently raises the technical bar through thoughtful code reviews?”

“Which engineers lead high-impact designs, and how were those received by peers?”

“Who actively supports customers or internal teams in Slack, Jira, or Google Docs?”

At Databricks, we’ve begun using internal LLM-powered workflows, built on models like Claude and our in-house Goose-based platform, together with MCP integrations. These workflows use indexed data across GitHub, Google Docs, Slack, and Jira to provide a continuous, evidence-backed view of individual performance. They can cluster contributions by project, highlight authored design docs and meaningful code reviews, trace technical discussions in Slack, and analyze PR activity.

The outputs are generated in markdown using prompts tailored for performance reviews, making them easy to drop into review documents or other tools. By default, managers receive ready-to-use summaries, but they can also ask the LLM clarifying questions when needed. For example, they might highlight specific design contributions or surface relevant Slack discussions. This removes the scramble of digging up links and notes during review season, saving at least two to three hours of data gathering per employee, and providing managers with consistent, review-ready inputs.

This shifts performance management from a static, manager-driven narrative to a more balanced process. Managers make better decisions because they see a fuller view of an engineer’s impact, not just the most visible or recent work, while engineers gain clearer visibility into how their contributions are recognized and valued.

As a recent Fast Company article noted, AI is beginning to transform performance reviews by helping managers gather insights across fragmented tools and systems. The result is a more complete and consistent view of contributions, even in fast-moving environments.

Even without these advanced integrations, one can leverage various other tools. Notebook LLM can ingest design docs, peer feedback, and self-reviews, enabling me to query and explore each engineer’s work more interactively. Instead of skimming dozens of documents, I could ask targeted questions and get clear summaries. ChatGPT further helped structure the reviews by turning rough notes into well-phrased evaluations. What initially felt nearly impossible to complete became not only manageable but efficient.

Without these tools, completing over ten reviews would have taken days or even weeks. With them, the process became faster, more accurate, and far more scalable.

Hiring and interview feedback

Hiring is one of the most impactful responsibilities for engineering managers. But with multiple candidates in the funnel and interviews often scheduled back to back, it can be difficult to capture clear, actionable insights before details fade. AI tools have made this part of the job significantly easier and more effective.

Transcription tools like Gemini make it possible to stay fully engaged in conversations instead of scrambling to take notes in real time. In many interviews, managers are seen typing continuously, appearing distracted or disengaged. Many managers have had to build note-taking as a core interview skill, but that is becoming less relevant. With transcripts available afterward, the focus shifts to deeper listening during the interview and more complete review afterward.

A more effective workflow combines annotated notes, interview transcripts, and tools like ChatGPT to generate structured feedback aligned with hiring rubrics. While Gemini offers automatic summaries, they may not capture the depth or nuance needed for complex decision-making. Using ChatGPT helps surface key examples, refine phrasing, and strike the right tone for hiring committee review.

What once required real-time multitasking and rushed feedback can now be handled with more clarity, consistency, and flexibility. Feedback can be submitted later in the day, without losing context or signal quality.

Execution and strategy

Execution and technical alignment are core to an engineering manager’s role. Yet building the necessary context to make sound decisions, review designs, and guide a team toward reliable delivery often demands time and attention that managers simply don’t have.

As team scope expands and projects touch more surface area, staying close to both technical depth and product direction becomes harder. This is especially true when managers are expected to uphold design quality, drive reliability, and track customer impact, often in systems beyond their own direct authorship. AI tools are increasingly helping managers scale this part of the job by accelerating context-building, enabling faster tradeoff decisions, and providing visibility across both local and adjacent systems.

This trend is not unique to a single company. According to LeadDev’s 2025 AI impact report, 98% of engineering teams now use AI tools, up from 61% just a year ago. The shift is already underway and the leverage is real.

Building technical judgment faster

Line managers are expected to uphold the same level of technical judgment as a strong Staff engineer, across areas they may not be directly hands-on with. That means reading and reviewing design docs, understanding product use cases, and identifying risks before they turn into problems.

AI tools make this faster and more tractable. I often use Claude to ramp up on unfamiliar systems by summarizing critical code paths, highlighting patterns, and answering questions like:

• “Where are the retry and fallback behaviors implemented?”
• “How does rate limiting apply across different traffic types?”
• “What’s the difference between the legacy and new admission control?”

AI tools help analyze design documents at scale. Instead of skimming a dozen 10-page specs, I can interactively extract tradeoffs, validate customer impact, and check for alignment with SLAs and reliability goals.

These tools help me stay technically grounded, so I can uphold quality without bottlenecking on tribal knowledge or waiting for walkthroughs.

Communication and influence

Engineering managers spend a large part of their day communicating: aligning stakeholders, writing updates, sharing proposals, and providing feedback. But unlike meetings or live conversations, most of this communication now happens asynchronously through Slack, email, and docs. While this has made teams faster, it has also introduced new challenges.

The risk of misunderstood messages

One of the biggest pain points in async communication is that intent often gets lost. You might write a Slack message with the best of intentions, trying to be encouraging, respectful, or collaborative, but it can come across as cold, vague, or even dismissive. The reverse happens too. I’ve misread messages from teammates that felt abrupt, only to realize later the tone wasn’t intended that way.

In real-time conversations, we get more chances to adjust. There is body language, emotional cues, and the opportunity to clarify if something lands poorly. Many times, after struggling with a message, we just say, “let’s meet over a call,” and things get resolved in five minutes. That gap between intention and perception is something managers face every day.

LLMs have become a powerful tool for reducing that friction. When writing to senior leaders, navigating sensitive conversations, or just trying to be clear without sounding robotic, I often use ChatGPT to:

• Refine tone to match the intended spirit, whether it is friendly, direct, or empathetic
• Rephrase messages for specific audiences, such as executives or peer leads
• Add clarity and structure to ideas that initially came out rough
• Sense-check my language before sending high-stakes messages

It is not about outsourcing the message. It is about making sure what I meant to say is actually what is heard.

Proposal writing

Another area where AI has helped is proposal writing. In the past, starting a new initiative meant carving out a large chunk of time just to structure your thoughts. The blank page was intimidating. Even when you had a solid idea, turning it into something that could drive alignment and action took time and often delayed the start of meaningful work.

Recently, I wanted to kick off a design review committee. Instead of staring at an empty doc, I shared a few rough bullets and context with an LLM. In minutes, I had a clean, structured draft that outlined the goals, scope, process, and benefits. It was not final, but it gave me something concrete to iterate on and, more importantly, to share and socialize quickly.

This kind of leverage has made proposal writing more accessible. It encourages experimentation. Managers can move from idea to momentum faster, without getting stuck on formatting or phrasing. The barrier to starting something new is now much lower, and that opens the door to more bottom-up innovation.

The opportunity ahead

Engineering managers have always had to balance people, execution, and communication, often with limited time, fragmented information, and rising expectations. The job is not getting any simpler. But with the rise of practical and accessible AI tools, it no longer has to remain so manually intensive.

Across performance reviews, design decisions. and strategic communication, these tools are beginning to reduce overhead and amplify the impact of strong management. They help managers build context faster, write more clearly, and lead more effectively, without losing the human judgment that defines great leadership.

Even at the executive level, this shift is taking hold. Sundar Pichai recently told Google employees to “work smarter, not bigger,” encouraging teams to rely more on AI and less on traditional collaboration overhead. The message is clear: AI is not just for developers or data scientists. It is becoming a critical lever for technical leadership.

Developers got their AI moment. It is time engineering managers had theirs too.