
Hey! We are over at the other site too - groxio.bsky.social

Beginner or pro, Author & expert Bruce Tate's tailored approach to Elixir training will help you think beyond routine code to beautifully designed software.

Hey! We are over at the other site too - groxio.bsky.social

Writing LinkedIn articles isn’t always about being the author. Sometimes the better move is putting on an editorial hat and using Claude to shape the research and ideas. #AIWriting #LinkedIn

Agentic prompting works better with the right context: enough information about the task, the goal, and how to find relevant details. Better context leads to more useful results. #AI #PromptEngineering

Bruce Tate wrote code to teach developers. Now he uses agents trained on public repositories, including work like his own. The Blog, Cold Sweat, asks who pays for that convenience, and what it means to benefit from it. grox.io/blog/70-cold-sweat

Context Matters! Paulo Valim talks about building context for your agent in the most recent Groxio Live. Video will be available for 24 hours then available to Groxio customers. youtu.be/1oUrlH99GSk #AIEngineering

Ask an AI to solve a problem without naming a language, and it usually reaches for Python or JavaScript. Even with niche questions, it falls back on what it knows best—not always what was asked. #AI #Programming

A tetromino can have the right shape and still appear in the wrong place. Rotate inside its local grid first, then move every point to the board location. The order protects the offset. grox.io/blog/61-moving-the-w… #ElixirLang

Branch first, verify before review. Put that in the repository guidance and the agent's work arrives as a pull request you can inspect, not a chat you have to reconstruct. grox.io/blog/59-git-discipli… #AIEngineering

Enter `[99, 97, 116]` in IEx and you may see `~c"cat"`. The value is still a list; only the display changed. Inspect it with `IO.inspect(value, charlists: :as_lists)`. grox.io/blog/58-character-da… #ElixirLang

Your coding agent picks Python. That’s popularity, not a recommendation. Bruce Tate explains agent language bias, accuracy benchmarks and choosing a language your agent can write well. Free live: Oct 22, noon ET / 16:00 UTC. grox.io/live #AIEngineering

Bruce (@redrapids) uses AI every day, from family meals to preparing questions for his doctors. The token price still leaves out costs borne by people who never see the invoice. His new guide starts to itemize them. #AIEngineering grox.io/blog/57-read-the-lab…

Lots happening with ExMex Conf 2026! The schedule is live, lightning talk submissions are open, and standard ticket pricing ends in one week. Plus, if ticket cost is keeping you from attending, community tickets are available to request through Oct 21. exmexconf.com

The code can be committed while the project state disappears. In AI work, prompts, plans, and guidance need a durable home too. grox.io/blog/52-three-source… #AIEngineering

A `MapSet` is not a list with nicer membership syntax. It is a set-shaped API over map-shaped machinery: keys do the work, values disappear. grox.io/blog/55-mapsets-a-se… #ElixirLang

TODAY! Bruce Tate Live: LLM Harness MCP Demo — Part 2. Bruce continues the September 17 demo and takes questions live. Thursday, Oct 8 · noon ET / 16:00 UTC. Register free: grox.io/live

Vibe coding fails when assumptions stay hidden across all three agent inputs: priors, evidence, and correction. Put versions, preferred paths, and fast feedback where the agent can check them. grox.io/blog/54-vibe-coding-…