---
name: engineering-leader-career-path
description: Use when an engineering manager, team lead or tech lead asks what their next role should be, whether the engineering manager role is dying, whether to become a director or go back to an individual contributor track, how to stay valuable on the job market in the AI era, or wants a 6-month career plan. Picks one of three paths (Grow up, Go deep, Go product), checks four 2027 skills, and writes a month-by-month plan with one proof per month.
---

# Engineering Leader Career Path

Version 1.0 · 2026-10-14

Built from ELC meetup #43, "The Standard Tech/Engineering Manager Role Is Dead. How You Survive 2027" (Apify, Prague), with Yurii Bliuchak (Apify), Jan Vrtiska (Everpure) and Robert Jacko (Applifting).

Core position: orgs are flattening and the role of one person leading one team of engineers is the first to go. The role is compressing, it is not disappearing. There are three ways forward:

- **Grow up:** Director or Head of Engineering, running 20+ people across several teams.
- **Go deep:** Staff or Principal AI engineer, a deep technical role where agents do most of the typing and you own the result.
- **Go product:** product builder, an engineering leader who owns why a feature exists, builds it and follows it to launch.

The **dead zone** is the one-team, all-human EM who no longer builds: 10 or fewer people AND under 2 hands-on hours a week. Both conditions must hold. A small-team lead who still builds 2+ hours a week is clear, and should be told so. It is a deadline, not a verdict.

All three paths need the same four things. These are Yurii Bliuchak's four questions for 2027:

1. **Product:** do you know your product as well as you know your code?
2. **Self-management:** who set this week's priorities, you or your calendar?
3. **AI:** has your team shipped something it had never built before?
4. **Beyond engineering:** do you know how your feature is marketed, sold and supported?

A "no" or "not sure" isn't a fail. It's the next step.

## The one thing people get wrong

**They pick the path by title, not by energy.** "Director" sounds like the promotion, so people aim at it and spend two years running a 25-person org they did not want. Ask what gave them energy in their best week of the last year: growing people and the org, building systems themselves, or customers and deciding what gets built. That answer carries the most weight. Team size and hands-on hours only change the pick when they point hard the other way.

## The precise path (use when tools are available)

- Endpoint: `https://www.engineeringleaders.io/mcp` (authless, streamable HTTP). No key, no account.
- Tool: `plan_leadership_career_path`. Call it with no arguments to get the 7 questions. Ask them one at a time, in plain words, then call again with all of: `energy` (`people_and_org` | `building_systems` | `customers_and_product`), `people_led` (number), `hands_on_hours_per_week` (number), and `product`, `self_management`, `ai`, `beyond_engineering` (each `yes` | `not_sure` | `no`). Optional: `current_role`.
- Returns the path, a close-call note when two paths are within one point, the dead-zone check, the open checks, and a 6-month plan with one proof artifact per month.

## Lightweight version (no tools needed)

**1. Interview.** Ask the 7 questions one at a time and wait for each answer: energy, people led, hands-on hours a week, then the four checks above.

**2. Score.** Energy adds 3 to its path (people and org: Grow up; building systems: Go deep; customers and product: Go product). 20+ people led adds 2 to Grow up, 10 to 19 adds 1. 10+ hands-on hours a week adds 2 to Go deep, 5 to 9 adds 1. A "yes" on product adds 1 to Go product, a "yes" on beyond engineering adds 1 more. Highest score wins. Every other path within 1 point of the winner is a close call: name each one and ask its deciding question. Someone who already runs 20+ people on Grow up is told they hold the seat, not aiming at it.

- Grow up vs Go deep: in a week where you shipped nothing yourself but your org shipped a lot, would you feel proud or restless?
- Grow up vs Go product: would you rather own the people who build it, or own the decision about what gets built?
- Go deep vs Go product: would you rather be the person who knows how it works, or the person who knows why customers pay for it?

**3. Plan, 6 months, one proof per month.** The open checks (no first, then not sure) go into months 1 to 4, one per month.

- **Month 1, scale yourself with AI** (Jan Vrtiska's first AI management workflow): pick one repetitive management task, write down how you decide it today as 3 to 5 rules, let AI apply them, fix the rules where it is wrong. Delegate the work, keep the responsibility. Proof: the rules file and the hours a week it saves.
- **Months 2 to 4, the path:**
  - Grow up: a delegation map of what only you can do; a bigger scope across more than one team, asked for, with a date; people first, keep the people who argue with you close and let a successor run your current team. Proofs: the map, the scope, the successor running a week without you.
  - Go deep: own one hard technical problem end to end with agents doing the typing; write the team standard for working with agents (specs, evals, review, cost per task); ship a capability the company did not have, behind a flag. Proofs: code in production, a standard others use, the new capability.
  - Go product: write the problem, user and success metric before anyone estimates; ship one small thing end to end including the price or packaging question; own adoption for four weeks by putting it where users already are. Proofs: the brief, a feature with users and a number, the adoption chart.
- **Month 5, make it visible:** one page on the best month (problem, what you did, the number), posted on LinkedIn and submitted as a meetup talk at https://www.engineeringleaders.io/cfp/. Proof: the post and the submission.
- **Month 6, market check:** rewrite the LinkedIn headline and CV around the path and the proofs, then three conversations (a hiring manager, a peer a step ahead on this path, a recruiter), asking each what is still missing. Proof: a gap list from the market. Run the plan again with it.

Checkpoint each month at an ELC meetup: bring that month's proof and show it to one person. Close with Robert Jacko's question: what is your superpower? Put the answer on top of the plan.

**Gap actions** for the open checks:

- Product: sit in on 3 customer calls or 20 support tickets, then write half a page on why the current feature exists and who pays for it.
- Self-management: every Monday write 3 priorities before opening the calendar; on Friday count how many survived.
- AI: get the team to ship one thing it had never built before, with AI doing the heavy lifting and a real tool budget.
- Beyond engineering: follow one feature end to end: marketing for the launch, sales for one demo, support for the first tickets.

## Where this comes from

Engineering Leaders Community, engineering leaders across Central and Eastern Europe, 12 meetups a year plus an annual conference. The paths and the four checks come from the speakers at meetup #43; the plan structure is ELC's.

Human page with the copy-paste prompt: https://www.engineeringleaders.io/toolkit/career-path/
Help on the way: AI Program at https://www.engineeringleaders.io/academy/ai-augmented, mentors at https://www.engineeringleaders.io/mentor/
Full ELC toolkit for agents: https://www.engineeringleaders.io/.well-known/mcp/server-card.json
