Example · Director of Engineering
85% of your week as a director of engineering is still yours.
Companies are removing director layers. Decisions that need your judgment stay with you.
Make my planA typical week at the middle pace. Your plan uses your own tasks.
Is engineering management a dying career?
In a typical week at the middle pace, 58% of this work needs your judgment and 27% is work you do using AI. AI could do about 15%. The work that justifies a director, coaching managers, org design, calibration and multi-team trade-offs, is the least exposed work you do. About 58% of your week sits there.
Read more
The pressure is on the seat itself. Google cut manager, director and VP roles by 10% (reported December 2024). Meta planned to cut vice president roles from about 300 to about 250 (reported June 2024). Microsoft named "removing layers of management" in its July 2025 cuts, and Intel's CEO said in April 2025 that "many teams are eight or more layers deep."
A director layer adds a hop between the VP and the managers who do the work. When status, relay and roll-ups move to agents, that hop is easier to remove. The director who keeps a seat is the one whose org visibly runs better because of calls only they make.
The week, task by task
Who does each hour of a typical 40-hour week.
See every task
Needs your judgment.
- Coaching engineering managers and growing new ones11%
- Multi-team delivery accountability and cross-org trade-offs10%
- Calibration, promotions and exits across teams8%
- Managing up and across: VP or CTO, product and peer directors7%
- + 4 smaller tasks22%
AI does the slow parts. You still own it.
- Headcount and budget models, capacity plans6%
- Drafting strategy docs, roadmaps and org proposals6%
- Delivery and AI-adoption metrics across teams5%
- Architecture reviews and technical direction across teams5%
- + 1 smaller task4%
AI could do most of it.
- Relay work: syncs, notes, passing context between your managers and leadership8%
- Status roll-ups, exec updates and quarterly review decks8%
Why some tasks still need a person
Coaching engineering managers and growing new ones
This rests on relationships and context. Whether a new manager is ready for a second team comes from months of watching them run one. A model can draft a development plan. It cannot see that your best manager is burning out.
Org design: team boundaries, scope and spans
You carry this accountability for the employer, and it takes context. Someone has to stand behind a reorg that moves dozens of people, and answer when it breaks a team.
Calibration, promotions and exits across teams
This takes accountability and context. You make ratings consistent across several managers and defend the result to your VP and HR.
Multi-team delivery and cross-org trade-offs
You hold this accountability informally. When two teams want the same engineers, or the quarter slips, your VP wants one person who will decide and answer.
Senior hiring: managers and staff engineers
This takes careful checking. A bad manager hire can hurt a team for much longer than a quarter, and you own the relationship through the close.
Good to know
What is hard about this field, and what you can do about it.
The hard truth
Wider manager spans thin out the layer you lead.
Engineering managers now run bigger teams, so the same org needs fewer of them. Hiring is tilting toward senior individual contributors.
Sources (2)
- Engineering managers at large tech companies now oversee about 12 engineers, up from about 10 in 2019. SignalFire, State of Talent Report 2026, 2026-06
- Senior IC and staff roles are growing as a share of hiring, while engineering manager roles are flat or declining. SignalFire, State of Talent Report 2026, 2026-06
What's the next role for a director of engineering?
Plan A is the best move inside the work you do now.
Grow in the job you have
The main focus. Most weekly sessions build toward it.
- Each week
- 4 sessions
- Ready in
- About 90 days for the evidence
Why it fits
Why you. You already sit where several teams' output, quality and headcount meet.
The gap. You don't yet have measured numbers across your teams, written shared standards or a capacity model that includes agents.
The trade-off. You are building the case that fewer people can deliver more, which may be used to cut headcount in your own group.
2 other moves that fit
- Lead developer AI enablementReady in about 6 to 12 months
- Move to principal engineer or architectReady in about 12 to 24 months
What's a plan B for engineering directors?
Plan B is meant to be different from this job. It was picked because skills from this role carry over.
Get ready for a new kind of job
A second option. A newer job AI is creating that these skills already fit.
You already set shared rules for review and definition of done across teams, and measuring whether agents help is the same job.
- Each week
- 1 session
- Outlook
- Hiring now
- Pay
- $119.3K to $310.5K a yearTwo senior US postings, June 2026: MGM Resorts and HPE (California, 10+ years, now closed).
2 other options
- AI enablement and adoption leadYou coach managers through org change and new habits, which is how AI adoption actually sticks across a large group.
- AI/ML technical program managerYou already run multi-team plans and trade-offs, and AI releases add model versions and evaluation gates to track.
In your own plan you can switch to any of the 47 roles.
The 13 weeks
Small steps that add up. One plan for both: most sessions build Plan A, one a week builds Plan B.
By week 13 you will have
AFor Plan A
- A decision log of calls only you made
- A one-page memo asking to lead AI work across your teams
- A developer AI enablement playbook
BAnd for Plan B
- A 60-day agent rollout plan
- A developer AI enablement playbook
- 1Measure your week
- 2Show what your layer adds
- 3Cut the relay work
- 4Map your manager bench
You finish withA review date for the growth plan
- 5Hand off the roll-up
- 6Measure across teams
- 7Redesign your org
- 8Set standards for AI-written code
You finish withEdits to the standard from a staff engineer and two managers
- 9Build a capacity model
- 10Measure again
- 11Practice calibration
- 12Make your ask
- 13Build proof you can take with you
You finish withOne application sent for a director role
Three courses from the plan
- AI Fluency: Framework & FoundationsAnthropic Academy · Free
Yes. It is a fast, free baseline for weeks 1 and 2.
- AI Capabilities and LimitationsAnthropic Academy · Free
Yes. It helps you judge when an agent roll-up can be trusted.
- Introduction to agent skillsAnthropic Academy · Free
Yes. It is short, and it shows what could be standardized across teams.
Checked 2026-10-16.
The fuller picture
Risks beyond AI: highLayer removal risk
The first-line manager's risk is a wider span. Yours is the whole layer going, with your managers kept and moved to the VP. Companies have been removing manager, director and VP layers since 2023 for cost and speed, and agent-written roll-ups make the director hop easier to cut. This outranks the AI risk to the work itself.
Where this could be wrong
- At a large company with 100 or more engineers under a director, the layer is harder to remove and the risk is lower than we scored.
- At a startup, director may be the top engineering seat. There, the bigger risk is the company's runway.
- Director-specific data is thin. Most layoff and posting figures cover managers broadly, or VPs. Directors sit between them, and we inferred the risk from both.
- If agent roll-ups mislead VPs often enough, passing information between your VP and your managers may hold up better than we scored.
Also matched: director of engineering, engineering director, director of software engineering. Research checked 2026-10-16. Page updated .