I'm Augustine: a husband, father, friend and leader. My work is mostly with people, so I do a lot of listening, asking questions and pulling good folks together around what matters. I can't let go of quality, discipline, accountability, resilience, good debate and all others that build great character.
This is the place where I think out loud on heart, mind and hands from personal and professional experiences. My writings (I call them drops) usually do not take more than 40 seconds, or at most 4 minutes, to read.
Now listening · June 12, 2026
Just started Peak — the book behind the famous "10,000 hours" idea, except Ericsson argues it was never really about the hours. It's whether the practice is deliberate: focused, uncomfortable, always just past what you can already do. Early chapters. I'll report back when I've finished.
31s · June 12, 2026
Finished the audiobook · April 29, 2026
I finished Morgan Housel's Same as Ever, and one line has been following me around: nothing worth pursuing is free.
His point is that we treat the cost of good things as a malfunction — the stress, the bureaucracy, the hard conversations, the tears — as if something has gone wrong. It hasn't. That's just the fee, and it's non-negotiable. The grass looks greener on the other side mostly because it's fertilized with someone else's hidden costs. Every team worth building I've been part of came with a price I didn't want to pay.
The cost isn't a sign you chose wrong — it's the receipt for something worth having. Decide what you're willing to pay before you decide what you want.
27s · June 11, 2026
I used to judge a tool by its demo. If the walkthrough was clean and the dashboards were pretty, I trusted it.
So I pushed hard for self-hosted PostHog to capture events across our mobile and web apps. The demo was beautiful. Under our real load, with our data and our constraints, it came apart — built for side projects, not production. We ripped it out, having paid for the lesson in weeks of engineering time.
A demo shows you the happy path on someone else's data. Evaluate every tool under your real load, with your real constraints — or you're just buying the marketing.
26s · June 10, 2026
The San Francisco leadership canon treats direct, depersonalised feedback as a universal good. I imported it wholesale and expected it to work.
It misfired. In the West African teams I lead, trust isn't assumed at the start of a conversation — it's built relationally, over time, through rituals the books never mention. Feedback that reads as refreshingly "candid" in San Francisco can land as a breach of that trust here.
Management advice ships with invisible cultural defaults. Before you adopt a Western playbook, ask what it assumes about trust — and whether that assumption holds where your team actually lives.
25s · June 9, 2026
Early on I believed that if the code was clean, the system would work. Tidy abstractions, good tests, green build — surely that was the job.
Then I built fintech infrastructure in West Africa. The people moving money at 11 PM on a Friday found every fault my test suite missed. Edge cases weren't edge cases — they were the cases. The code was clean. The system still broke.
Correctness lives in the messy reality your code meets, not in the code itself. Design for the degraded state and the hostile input first; elegance is what's left over.