Issue #643

Essential Reading For Engineering Leaders

Friday 22nd August’s issue is presented by Buf

Nike reduced the time to create new APIs from many months to days. They switched to a Protobuf-first approach and moved from a monorepo nightmare to versioned remote modules with the Buf Schema Registry.

Proper data architecture pays massive dividends:

— Alex MacCaw

tl;dr: (1) Set up strong scaffolding and type-safe context. (2) Codify conventions in .cursor/rules. (3) Load all relevant files—including types—into Cursor’s context. (4) Use planning mode on top-tier models (o3‑pro, Claude Opus 4, etc.) (5) Lint, typecheck, and test after each plan execution. (6) Embrace audio-first prompting. (7) Enjoy these last few years of human-led coding magic.

Leadership Management

— Wes Kao

tl;dr: “If your manager finds it difficult to communicate with you, there’s always friction, and you just don’t get along that well, it doesn’t matter if this conflict is due to their personality defects. I repeat: It doesn't matter if the tension is technically “their fault.” You have to find a way to deal with it, or work for someone else.”

CareerAdvice

— Scott Haines

tl;dr: Nike reduced the time to create new APIs from many months to days. They switched to a Protobuf-first approach and moved from a monorepo nightmare to versioned remote modules with the Buf Schema Registry. Proper data architecture pays massive dividends: (1) Backward compatibility guaranteed: Impossible to force breaking changes. (2) Type safety at compile time: Catch errors before production. (3) Decentralized collaboration: Teams work autonomously while sharing schemas.

Promoted by Buf

Management Data API

— Gergely Orosz, Elin Nilsson

tl;dr: This post covers: (1) Most-mentioned tools. (2) Project management. (3) Communication and collaboration. (4) Databases and data stores. (5) Backend infrastructure. (6) Forks of popular open-source projects.

Leadership Management

“Testing leads to failure, and failure leads to understanding.”

— Burt Rutan

— Jerome Bellegarda, Glen Xiao

tl;dr: “This post investigates the benefits and limitations of prompt engineering in two instances of AI-assisted onboarding relying on large language model (LLM) technology. Of particular interest is how truthful (and therefore reliable) LLM-generated answers turn out to be in the context of Etsy-specific question answering. Among other insights, we find that asking the LLM to identify specific source snippets is a good way to flag potential hallucinations.”

CaseStudies AI

— Chris Kelly

tl;dr: Chris Kelly from Augment Code breaks down why the "AI will replace developers" hype misses the mark. Why professional engineers are the last to adopt AI tools, the critical difference between vibe coding and production-ready development, and practical tips for working with AI that treats it like what it is—a machine, not magic.

Promoted by Augment Code

Video AI

— Matthias Döpmann

tl;dr: “In Software Engineering, we easily set targets around metrics and processes which have a constant behavior, regardless of whether we have 50, 300 or 50,000 lines of code. At the same time, we do not pay close attention to those metrics that change as complexity grows, and only realize their impact when the pain is obvious.”

Leadership Management

— Dennis Schubert

tl;dr: “I like data. That’s about as surprising as the sun rising in the morning. A big part of that is visualizing various aspects of my life. I consider myself lucky enough to be able to travel quite a bit, so I always liked having a visual history of all the places I’ve been - and I’m sure lots of people can relate to that. I used to be quite happy with the Google Maps Location Timeline, but I stopped that a while ago for obvious privacy reasons. So I decided to build my own, and that’s what I’ll be talking about here.”

Geolocation

tl;dr: “Airport trips require careful consideration and strategy, often posing a more complex experience for drivers than city trips. As a result, many of Uber’s typical city marketplace models and algorithms are no longer applicable at airports, requiring novel solutions more tailored to airport dynamics.”

ML

Prompt > Planet

Hand Drawn by Manu

Product Engineer, Fraud Prevention: Clerk is hiring a mission-driven Software Engineer for our Fraud team. You’ll build systems to detect and prevent fraud, protect users, and partner cross-functionally to stay ahead of bad actors—balancing quick wins with long-term vision.

Epicenter: Local-first, OS apps.

Vibe Coding: Tips and tricks.

POML: Prompt orchestration markup language.

Sim: OS AI agent workflow builder.

Try: Fresh directories for every vibe.


How did you like this issue of Pointer?

1 = Didn't enjoy it all // 5 = Really enjoyed it
1  |  2  |  3  |  4  |  5

Login or Subscribe to participate in polls.