AI can write code for us now. 🤖 But that makes programming fundamentals more important, not less. Generating code is becoming cheap. Knowing enough to question what AI gives you isn't. 👩🏾‍💻 675K+ learners later, I still believe in the basics: linkedin.com/learning/progra…

I think we’re going to outgrow the “prompt library.” I’ve started adding metadata/frontmatter to prompts: ID, owner, version, inputs, tags. 🧩 Once prompts run real workflows, they’re not just text anymore. They’re engineering artifacts with a lifecycle. #AIEngineering

Instead of suffering through another week of meetings that could’ve been an email 😅, add some P.O.W.E.R.: Purpose Outcomes What’s in it for them Engagement Roles + responsibilities Better meetings start before the invite gets sent. linkedin.com/learning/tech-c…

This week's working theory: AI may speed up engineering faster than it speeds up the rest of product development. Several conversations reinforced a pattern: improving SDLC flow is valuable, but the next bottleneck shows up somewhere else. Still testing that hypothesis. 💭

I've been reworking my AI workflow instead of chasing the newest model. More deliberate routing between skills, local models, GitHub Copilot, and other assistants has reduced token usage without reducing how much engineering work I'm getting done. 🛠️ annycedavis.com/2026/09/toke…

I've been digging into engineering metrics in teams that use AI heavily. One weird finding: a PR with ~50 review comments looked risky until we opened it. Most of the comments came from AI. I think we're going to need AI-aware eng metrics, not just AI adoption metrics. 🔬

Local AI finally found its lane for me. 🤖 Aider + Ollama handles straightforward coding work where privacy matters: refactors, cleanup, repetitive changes. Not a replacement for stronger cloud models. Just another tier in my toolbox. 🛠️ #AIEngineering

This week I kept running into decisions where both sides were reasonable: consistency vs. latency, governance vs. speed, planning detail vs. flexibility. AI can help explore the options. Judgment is deciding which tradeoff to make. 🛠️ #SoftwareEngineering

I built a commitment matcher using stop words + token overlap instead of embeddings. 🔧 Deterministic, explainable, and fast enough for a CLI. A good reminder that not every fuzzy matching problem needs embeddings. Sometimes the boring algorithm is the right answer.

I keep finding places where the answer isn't "let the AI do more." 🗂️ My file organizer got better when I assumed most files were already where they belonged. Rename freely. Move rarely. That one constraint cut my review time in half.

🤖 I stopped defaulting to the same AI for every coding task. I built a skill that scores the task, privacy, scope, cost + available models, then recommends the right tool/model. I’m starting to think of this as AI workload routing. #AIEngineering #SoftwareEngineering

I'm becoming convinced that "% of engineers using AI" is mostly a vanity metric. 📊 Someone using Copilot twice a week and someone rebuilding their workflow around AI both count as "adopted." The number goes up. But did anything actually get better?

For years I thought great work was enough. It isn't. The engineers and leaders who stand out make their thinking visible. That's the thread connecting my work on Invisible Signals™, AI, and leadership. My latest newsletter ==> open.substack.com/pub/brwngr… #Engineering #Leadership

🚨 No wonder nobody is calling. You spent years building expertise, but your resume, interview answers, and accomplishments send weak signals about ownership and impact. Capability matters. Signal quality matters too. 🔗 annycedavis.com/2026/06/invi… #Hiring #Leadership #Tech

What signal are you sending? Today I'm open-sourcing Invisible Signals™—a project focused on helping engineers and technical leaders understand the signals they send through resumes, interviews, leadership communication, and career growth. 🚀 invisiblesignals.dev

Most leaders are already using AI—a lot. 😅 They’re just not sure what “good” looks like. So they get more output… not better decisions. AI isn’t the constraint. Cognitive drag is. 🤷🏾‍♀️ More: annycedavis.com/2026/04/ai-a…

The biggest slowdowns on engineering teams rarely come from lack of talent. They come from friction we’ve normalized. Jetpack Compose reduced more of that friction than I expected. Here’s what changed for me: annycedavis.com/2026/02/jetp… #EngLeadership #AndroidDev #JetpackCompose

#Androidified myself. I'm just a traveling cowboy! 🤠 Create your own and watch the latest episode of #TheAndroidShow. android-developers.googleblo…