Carlos Hernández — Engineer & maker, San Salvador

I build production LLM systems — and small tools for my own life.

Eight years building production cloud systems — and the engineering organizations behind them — at EsePlus / Alilo, where I now lead AI Platform Engineering. I treat LLMs the way I'd treat any platform: observable, version-controlled, costed, deletable. Also DevOps Engineer at Deckers Outdoor, and finishing a Master's thesis on comparative NLP for transaction classification.

2026 · May

Writing field notes on production patterns for LLM applications. Benchmarking classical NLP baselines against transformers for the thesis — surprised by how close they're landing.

I Production work

Alilo AI Suite · LXD AI + Alibot In production

Built at EsePlus / Alilo · Head of AI Platform Engineering

Two AI engines I designed and own end-to-end, exposed as agnostic APIs for the product team to consume and compose into experiences. Alilo LXD AI auto-generates corporate microlearning experiences from uploaded documents — async pipeline coordinating document analysis, content generation, and activity design. Alibot is the multi-tenant RAG conversational backend — hybrid retrieval, Matryoshka embeddings, Google ADK agent with multimodal tool calling, defense-in-depth safety with automated adversarial regression.

Highlights: 95% → 99.5% availability uplift · 20%+ cloud cost reduction · ~90% token savings on validation retries · sub-3s p50 latency · zero jailbreaks in adversarial testing.

h14z Independent

Consulting on AI/LLM infrastructure for early-stage teams

Strategic advisory for mid-to-large teams running real AI in production. Privacy-first LLM systems, FinOps, and cloud architecture — no SaaS, no vendor lock-in.

Comparative NLP for transaction classification Research

Master's thesis · Universitat de Barcelona (OBS) · defense Sep 2026

Comparative evaluation of four NLP approaches for automatic bank-transaction categorization — from classical baselines (TF-IDF + SVM, XGBoost) and embedding-based neural networks to transformer fine-tuning (BETO / DistilBERT). Full evaluation pipeline measuring precision, recall, F1, and inference latency across 68k+ labeled transactions.

II Workshop

Small tools I build for my own life. Each one started because something annoyed me enough to write code about it.

Cadence

A local-first planner that marries GTD with time blocking. Built because every productivity app either fights your method or assumes you don't have one yet. Works offline, syncs across devices via Realtime, ⌘K does everything.

Next.js · Supabase · TypeScript · cadence.carloshdez.com↗

Fitness Dashboard

A personal training and body-composition tracker — split routine, recomposition goals, observable progress. The dashboard I wanted but couldn't find as an app.

React · Tailwind · personal

III Field notes

May '26 Prompts deserve migrations, not vibes EN
9 min

Schema migrations have versioning, rollback, and a paper trail. Most prompts in production don't. What changed for me when I started treating them the same way, and where it broke first.

Apr '26 TF-IDF beat my transformer. I'm not surprised. EN
10 min

Halfway into a thesis benchmark on 68k bank transactions, the classical baseline is uncomfortably close to BETO. The catch isn't accuracy. It's what you pay per inference, multiplied by traffic you actually have.

Mar '26 What "observable" actually means for an LLM app EN
7 min

Infrastructure traces tell you the request landed. Quality metrics tell you the answer was wrong. Until both share an ID, you're guessing. A pattern set from running this on call, and the one thing I'd build first if I had to start over.

Feb '26 My IDE finally feels like a coworker EN
8 min

"Tool calling with extra steps" is the wrong framing. The shift is smaller and weirder: the editor reads our Jira, our repo, our traces, and starts answering the questions I used to walk over to a teammate for.

IV Speaking

Cursor + MCP — when your IDE knows your stack

Internal tech talk · Recording available

2026.02

V About

I'm a Senior AI Platform Engineer in El Salvador. I work from the same desk where I started in cloud infrastructure eight years ago. The systems have grown, the teams behind them too — but the principles haven't. Only the layer I apply them to.

I care about systems you can operate, not just demo. About prompts that fail loudly. About observability that survives the third on-call rotation. About cost lines that don't blow up the first time a feature gets traction.

Outside of paid work, I build small tools for my own life — usually with the same disciplines, because I don't know how to build software any other way.