Eduardo
García-Garzón
Head of AI · Shakers · Madrid
Hi, I'm Eduardo. I lead AI at Shakers, a flexible-talent marketplace in Madrid where my team ships the matching, conversational, and agentic systems behind a market of 15,000 freelancers and 450+ companies — production grade, under EU AI Act constraints.
We're 1 of 12 Spanish companies selected by Spain's Ministry of Digital Transformation (DG IA) for the EU AI Act regulatory sandbox, and Shakers' matchmaking system is in the OECD.AI Observatory catalog. Background in applied statistics — psychometrics, multivariate methods, measurement theory. I bring that lens to LLM evaluation, because what most people call “eval” is a statistics problem in disguise.
I write here about production AI, evaluation, and EU regulation from the operator seat. Outside Shakers, I speak at industry events when the topic is one I can speak about honestly.
Want to talk? Book a 30-min chat.
Writing
- An agent acts. A wrapper answers.July 24, 2026
Gartner says 40% of enterprise apps will embed agents this year, and that most vendors are faking it. Both are true because 'agent' stopped being a technical word.
- You can't price the work by the seatJuly 5, 2026
Enterprises burned their 2026 AI budgets in months. The token bill made pricing a strategic question — and the per-seat model is now a subsidy.
- Europe funded the layer that gets cheaper without itJune 22, 2026
On 19 June the Commission funded a sovereign frontier model. That is the one layer that commoditizes itself. The value already moved two layers up.
- The company is the productJune 5, 2026
Sequoia's thesis on the next trillion-dollar company says software is no longer the unit of sale. The work is. Headless is the architectural shape of the bet.
- The line the AI Act just drew through HRMay 21, 2026
The Commission's draft Annex III guidelines published today put two-sided talent platforms squarely inside the high-risk perimeter. The same engine, surfaced to the candidate, sits outside it. The verb is what changed.
- $5.5 billion to do the work consultancies do todayMay 13, 2026
On 4 May, OpenAI and Anthropic announced enterprise services joint ventures on the same day. Not the model. Not the API. The work. The middleman is the casualty.
- You can outsource thinking. You can't outsource understanding.May 4, 2026
Karpathy reframed LLMs as statistical distillates of text, not animals that learn the world. The shift from vibe coding to agent engineering is what production has been waiting for.
- 151,700 AI professionals in Spain. 6,700 who build.April 20, 2026
El País counted us. The handful who actually train models are doing something different than the rest of the AI workforce — and what we're doing is changing in a way the headline misses.
- Speed without alignment is noiseApril 6, 2026
Notes from a ProductTank Madrid we hosted at Shakers, on what changes when AI enters a product team — and what doesn't.
- Efficiency is the floor, not the ceilingMarch 23, 2026
PwC measured 1,217 executives across 25 sectors. 20% of companies capture 74% of the economic value of AI. The split is not budget, not stack — it is whether the company uses AI to do what it already does, or to sell what it could not before.
- Define capabilities before you pick the vendorMarch 9, 2026
Dorsey from Block and Botha from Sequoia published a new operating model for AI-native organizations. It looks like architecture, but it is really a strategy for not letting your roadmap become the limit on what your AI can do.
- In AI, closed code does not protect you. It isolates you.February 23, 2026
Anthropic's Claude Code leaked 512,000 lines via an npm source map while OpenAI shipped Codex CLI as Apache 2.0 from day one. The lesson is not about open source ideology. It is about where defensibility actually lives in this cycle.