RSSAmplifier

Blog

George Guimarães.

AI and Agentic Engineering in Elixir. Linking processes and people.

georgeguimaraes.comRSS feed ↗15 posts

Latest posts

You're a Dumb Security Leader if You Mandate Password Rotation

NIST, Microsoft, and the FTC all agree: forced password rotation makes you less secure.

One GitHub Account Forever. Ditch Your Company Policy.

Your GitHub account is yours, not your company's. Here's how to keep your contribution graph intact across jobs.

Hallmark: detect LLM hallucinations locally in Elixir

Score whether an LLM's output follows from your source material or if it's hallucinated. Runs locally in Elixir, no API calls.

Your Test Suite Should Hit the LLM, Stop Mocking It

Your test suite should call the LLM API. For real. Not a mock, not a recorded cassette, not a fake that returns canned JSON.

Arcana: Embeddable RAG for Elixir/Phoenix

Everything I've learned building RAG systems, packaged into one Elixir library. Hybrid search, agentic pipelines, GraphRAG, all running on your existing Ecto Repo and pgvector.

Stephen: ColBERT-Style Neural Retrieval for Elixir

ColBERT-style late interaction retrieval for Elixir. Per-token embeddings and MaxSim scoring for more precise search than single-vector approaches.

Vibe, Agentic, Organic: The Three Ways to Code in 2026

Everyone's an AI-assisted developer now. But the range of what that means is absurd.

Elixir/BEAM Doesn't Solve Everything for AI Agents. Addressing the Criticisms.

Addressing the three most valid criticisms of using Elixir and the BEAM for AI agents.

Alike: Semantic Similarity Testing with the Wave Operator

Test LLM outputs by meaning, not by characters. A wave operator for semantic similarity assertions in Elixir, powered by local embeddings.

Your Agent Framework Is Just a Bad Clone of Elixir: Concurrency Lessons from Telecom to AI

Python and JavaScript/TypeScript AI frameworks are reinventing what telecom solved in 1986. What 40 years of production-grade concurrency teaches us about building AI agents.

Live Coding Interviews Are Broken for Everyone. And Not Because of AI

Or why I really hate this shit...

The Boring URL Shortener: System Design Interviews on the BEAM

The classic system design interview opener: design a URL shortener. Each follow-up is supposed to be harder than the last. On the BEAM, they're boring.

Full Stack AI with Elixir: San Francisco Global Elixir Meetup Talk

My talk at the San Francisco Global Elixir Meetup on building full-stack AI solutions in Elixir: multiple chains, in-memory RAG, MCP clients, and neural networks in your supervision tree.

Elixir Mentor Podcast: Forecasting, AI Agents, and the BEAM

I joined the Elixir Mentor podcast to talk about Soothsayer, why the BEAM is the best runtime for AI workflows, and agentic commerce.

Dependabot and Phoenix 1.8.x's CodeReloader Don't Work Together

Unless you do this trick.