I make your team
deliver faster
by integrating AI into their processes

Freelance AI Engineer embedded in your team. I automate workflows, from product requirements definition to the implementation of agents and RAG knowledge bases. From idea to deployment.

Book a call From €600/day · EU VAT-registered
Focus RAG · Agents · LLM Ops
Stack Python · TS · MCP
Mode Remote EU · CET
Based in Madrid, Spain
Dario Herrera
Shipped
3 AI products
Experience
10+ years
Delivery experience across
OrangeTelefónicaAccionaAvangridGT MotiveHolcim
European Union

EU VAT-Registered Operator (B2B)

Simplified invoicing and hiring for European Union companies under the reverse charge mechanism (VAT Reverse Charge).

How I work

From Discovery
to production.

No slideware, no endless discovery. I embed with your team and take an AI automation from first call to a working, monitored system in production. A single owner for the full cycle — me.

Discovery

I sit with your team and your real data —the messy kind— to map how you work today and where the bottlenecks are. I don't start from a ticket, I start from the business problem.

  • Analysis of context, team, and current workflows
  • Detection of bottlenecks and AI opportunities
Deliverable

Improvement report + prioritized implementation roadmap

Architecture

I design the system end-to-end: what platform and tools to use, what processes to automate, what agents are needed, and where the knowledge comes from. I define integrations with your sources —Confluence, Notion, Qdrant— and the guardrails that keep output accurate and reliable.

  • Platform, tools, and processes to automate
  • Agent & knowledge base design (Confluence, Notion, Qdrant)
  • LLM selection & trade-offs (cost, accuracy, latency)
Deliverable

Solution architecture + evaluation and observability plan

Build & ship

I build the system and deploy it fast, with automated evaluations, monitoring, and human review where errors are costly.

  • Production-grade delivery
  • Integrations: CRM, APIs, voice
Deliverable

Production system + testing and hardening

Handover

I level up your team so the system keeps growing when I leave: documentation, work patterns, and an agentic workflow that becomes yours.

  • Knowledge transfer and training
  • Operational documentation (runbooks, AGENTS.md)
Option

Monthly retainer option for continuous evolution

Technology

The stack,
by layer.

Not a logo wall. The tools I actually ship with, organised the way I build a system — from model choice down to the ops that keep it reliable in production.

01 — Intelligence

AI & LLMs

Model selection by cost, accuracy & latency.

Claude Claude Google Gemini Gemini OpenAI OpenAI Ollama Ollama
02 — Orchestration

Agents

Multi-agent design, tools & context engineering.

LangChain LangChain / LangGraph n8n n8n Claude Claude Managed Agents OpenClaw Openclaw NousResearch Hermes
03 — Retrieval

RAG & Data

Retrieval over messy, real-world documents.

Pc Pinecone PostgreSQL pgvector / Weaviate PD PDF / doc extraction Em Embeddings
04 — Build & ship

Engineering

Production web apps, APIs & infrastructure.

Python Python TypeScript TypeScript Next.js Next.js M MCP CC Claude Code Docker Docker Supabase Supabase
05 — Reliability

LLM Ops

What keeps a non-deterministic system trustworthy.

Ev Evals Ob Observability Gd Guardrails HL Human-in-the-loop