← Essays

2 AI Tools I Use Every Week as an Engineering Leader

Everyone's talking about using AI to code or write. But very few are getting actual leverage from it.

Time is my constraint. Precision is my advantage. Here are my tools and workflows that actually give me “10x” developer productivity.

They actually save me time, make me smarter, and unblock real work across engineering and leadership.

1. CursorAI — The only IDE I’d consider paying for, after IntelliJ

Cursor isn’t just “Copilot inside VSCode.” It’s become my first stop for navigating complex dev work fast.

Here’s how I actually use it:

a. Building complex data warehouse queries

When I need to join multiple data sources across fragmented schemas, Cursor cuts through the noise. I describe what I want — the outcome, the filters, the edge cases — and it generates a solid starting point, complete with join logic, summaries, and optimizations.

b. Scaffolding new projects or services

I routinely need to spin up scripts, small APIs, or internal tools. I feed Cursor the requirements, architecture decisions, and constraints, and it scaffolds the code — tests included — in minutes. No boilerplate fatigue, no setup drag.

c. Fixing dev environment conflicts

Last week, I hit a dependency hell scenario with mismatched SDK versions across services. I dumped the stack trace into Cursor and YOLO’d (a real feature in Cursor) a resolution path that would’ve taken me hours of debugging. It flagged the root conflicts, suggested potential solutions, and downloaded and tried solutions step by step until resolution.

Cursor is no longer just a coding aid — it’s a debugging assistant, data analyst, and systems designer rolled into one.

2. ChatGPT — My Strategic Thought Partner (Not Just a Writing Assistant)

Most engineers still treat ChatGPT like a fancy autocomplete tool.

They’re missing the point.

I use GPT-4 as my thinking accelerator — especially when I’m strategizing under complexity and ambiguity.

Here’s how:

a. Strategy under constraints

In real engineering orgs, you don’t get ideal conditions. Limited bandwidth, messy ownership, competing goals — I describe the constraints and let GPT surface multiple viable paths. It helps me escape tunnel vision and generate higher-quality decisions, faster.

b. Tailored document templates

When I need to roll out a new testing initiative, prepare a quality review doc, or structure a postmortem — I describe the use case, audience, and tone. ChatGPT gives me a draft template I can adapt in 10 minutes. I spend my time on content, not formatting.

c. Alternative frameworks and mental models

I’ll prompt it with things like:

• “What are 3 alternatives to RACI for clarifying ownership?”

• “How can I map team accountability beyond service boundaries?”

• “What mental models help prioritize testing in complex systems?”

It surfaces fresh ways of thinking I wouldn’t find in the usual Slack echo chamber or PM handbooks.

One real example: I was able to blend RAPID (for decision-making), Jobs-To-Be-Done (for stakeholder alignment), and a risk-weighted testing framework — all from a GPT session that turned a vague idea into a structured plan.

Why This Matters

The biggest shift in engineering leadership isn’t that AI can code.

It’s that AI can think with you.

It gives you leverage:

• To move faster when the team is stuck.

• To sharpen your ideas before they hit a doc.

• To build systems that are resilient, not reactive.

If you’re not pairing with tools like Cursor and ChatGPT, you’re wasting hours on low-leverage work — and leaving better strategy, stronger execution, and deeper clarity on the table.

Don’t wait for a “perfect AI workflow.” Start small. Use one tool on one task this week. Observe the time saved and the clarity gained. Then go bigger.