# Salvador F. Criado

> Salvador F. Criado — AI Engineer and Platform Engineer in Granada (Spain), remote worldwide. LLM applications, RAG, agents, real-time voice, vLLM serving, AWS · Azure · Terraform · Kubernetes.

Salvador F. Criado is an AI & Platform Engineer based in Granada, Spain, working remotely worldwide. Eight years shipping production systems — event-driven pipelines, real-time telemetry, time-series at volume, and the Terraform and Kubernetes underneath them. Current focus: applied AI end to end — LLM applications, agentic systems, retrieval-augmented generation and real-time voice. Also covers the full local-model deployment cycle: quantisation, serving with vLLM and NVIDIA Triton, benchmarking and evaluation.

Positioning note: a versatile senior engineer with a strong systems foundation, with applied AI as the most recent layer of the stack — not an AI-only specialist.

Availability: Open to permanent, fully remote senior / staff roles. Remote worldwide, UTC+1, based in Granada, Spain. Contact: mail@salvadorfcriado.com.

## Pages
- [Home](https://salvadorfcriado.com/): who he is, selected work, experience, and how to make contact.
- [Field notes](https://salvadorfcriado.com/blog/): technical writing on production LLM systems.

## Topics
- [RAG](https://salvadorfcriado.com/blog/tags/rag/): Retrieval-augmented generation in production: what actually decides whether the model gets the right context, and what it costs to fix.
- [Search & Retrieval](https://salvadorfcriado.com/blog/tags/search-retrieval/): Recall, ranking and rerankers — telling a retrieval failure from a ranking failure, and the order to attack them in.
- [Evaluation & Monitoring](https://salvadorfcriado.com/blog/tags/evaluation/): Measuring an LLM system instead of arguing about it: evals, tracing, regression detection, and what to put on a dashboard.
- [LLM Serving](https://salvadorfcriado.com/blog/tags/llm-serving/): Running the model yourself — vLLM, batching, throughput and latency budgets, and the arithmetic behind the surprises.
- [LLMOps](https://salvadorfcriado.com/blog/tags/llmops/): The operational half: pipelines, reindexing, release and rollback, and the jobs that quietly decide whether answers stay true.
- [Data Engineering](https://salvadorfcriado.com/blog/tags/data-engineering/): The data side of a GenAI system — ingestion, freshness, and the scheduled work nobody wrote down and nobody put on call.
- [Regulation & Governance](https://salvadorfcriado.com/blog/tags/governance/): What compliance actually demands of an AI system, on what timeline, and which engineering decisions it quietly forces.

## Articles
- [The half of your GenAI system nobody put on call](https://salvadorfcriado.com/blog/the-half-nobody-put-on-call/) — 2026-09-10, tags: LLMOps, Data Engineering, RAG. Nobody's RAG system breaks — it quietly stops being right. The jobs nobody wrote down, three mechanisms that make them safe, and what to put on a dashboard.
- [Temperature zero is not deterministic](https://salvadorfcriado.com/blog/temperature-zero-is-not-deterministic/) — 2026-09-03, tags: LLM Serving, Evaluation & Monitoring, Regulation & Governance. Same prompt, same model, different answer. The mechanism behind non-determinism, the arithmetic that kills agent pilots, and what to do about it.
- [Your RAG doesn't have a retrieval problem. It has a ranking problem.](https://salvadorfcriado.com/blog/rag-ranking-not-retrieval/) — 2026-08-27, tags: RAG, Search & Retrieval, Evaluation & Monitoring. The document was in the index and the answer was still wrong. How to tell a retrieval failure from a ranking failure, what to fix first, and what each costs.
- [The gates are the product, not the orchestration](https://salvadorfcriado.com/blog/gates-not-orchestration/) — 2026-08-25, tags: LLMOps, Evaluation & Monitoring. The orchestration in an AI content pipeline is the boring part. What decides whether text ships is a set of model-free functions with an exit code.

## Optional
- LinkedIn: https://linkedin.com/in/salvadorfcriado
- GitHub: https://github.com/salvadorfcriado
- Email: mail@salvadorfcriado.com
- Feed: https://salvadorfcriado.com/rss.xml