Essays

The full archive.

19 essays. Start here if you want the argument first.

Start here

Aug 2026
AI for engineering managers without magic thinking

SpotDraft’s agentic SDLC is not a tool rollout. It is a designed loop from requirement to merge-ready PR, with agent-ready codebases, orchestration, and humans still owning the call.

ai engineering management developer productivity
Mar 2026
The engineering manager's second architecture

Software architecture is only half the job. The operating system around planning, ownership, review, and decisions is what determines whether a team can actually use that architecture.

engineering management execution architecture
Mar 2026
Mentoring engineers by giving them real ownership

Engineers grow when managers widen the decision space — not when they hand out harder tickets while still making every important call.

mentoring engineering management engineering leadership
Mar 2026
Stay close to the system without becoming the critical path

Technical depth still matters for engineering managers. The goal is leverage — better platforms, better engineers, and better judgment — not remaining the person who ships the hardest work.

engineering management architecture career
Mar 2026
Writing as a force multiplier for engineering leaders

Writing is leverage. It sharpens thinking, moves context out of one person’s head, and compounds in public the same way internal platforms compound inside a company.

writing career engineering leadership
Mar 2026
Platform thinking for smaller teams

You do not need a platform org to practice platform thinking. You need a habit of turning repeated pain into reusable defaults before the one-offs become the architecture.

platform engineering developer productivity architecture
Mar 2026
Better 1 on 1s for engineering managers

A 1:1 is working when it creates information a dashboard cannot, and both people leave with a next move that is actually owned.

engineering management mentoring
Mar 2026
Building trust in distributed engineering teams

Trust in distributed teams is predictability: written decisions, obvious ownership, and defaults that work when the author is not in the room.

engineering leadership execution
Mar 2026
Developer velocity without metric theater

Velocity improves when you diagnose the system — missing platforms, review bottlenecks, unclear specs — not when you celebrate the numbers that are easiest to count.

developer productivity engineering management execution
Feb 2026
Effective delegation for engineering leaders

Delegation is transferring a decision space with context and checkpoints — not moving a task off your plate while keeping the real call.

engineering leadership mentoring
Feb 2026
How to run better incident reviews

An incident review should change the operating system — ownership, observability, follow-through — not produce a document after a bad day.

reliability execution
Feb 2026
How to scale engineering decisions as teams grow

Decision speed dies when ownership is implicit. Growing teams need visible decision paths, not more people copied for safety.

engineering leadership execution
Feb 2026
Lessons from building in legal tech

Legal tech rewards precision, reusable workflow design, and enterprise integrations that do not become a new one-off for every customer.

integrations architecture
Feb 2026
Manager and staff engineer partnerships that actually work

The manager and staff-plus partnership works when technical direction and operating design are explicit, shared, and aimed at leverage rather than overlapping heroics.

engineering management architecture
Feb 2026
Meeting hygiene for fast moving teams

Calendars are an operating-system signal. Kill meetings that survive on habit; replace the ones that were covering for a missing written system.

execution engineering management
Feb 2026
Platform roadmaps that engineers trust

Engineers trust platform work when it removes a tax they already pay — not when the roadmap sounds mature and stays disconnected from the paved road.

platform engineering developer experience developer productivity
Jan 2026
Runbooks that improve delivery instead of collecting dust

A runbook is a default path for work that recurs under stress. If a reasonably new owner cannot use it, it is documentation theater.

delivery execution
Jan 2026
Skip levels that surface real information

Skip-levels are for texture the management chain filters out. Treat patterns, not gossip, and do not turn them into a shadow review.

engineering leadership mentoring
Jan 2026
The best career moat for engineers is judgment

Skills expire. Judgment about tradeoffs, systems, people, and timing is what makes technical depth useful as you take on more leverage.

career engineering leadership