Jorge Valverde-Rebaza

AI Researcher & Strategist | PhD in Computer Science | LLMs & Generative AI | Bridging research & real-world solutions in business, education & government

[Jorge Valverde-Rebaza @jvalverr]

✅ Data & model versioning
✅ Metadata tracking
✅ Automated pipelines
✅ Consistent training-serving
Reproducibility isn’t red tape, it’s strategy. That’s how you go from experiment to impact.

[Jorge Valverde-Rebaza @jvalverr]

If you can’t reproduce it, you can’t scale it.
Reproducibility isn’t bureaucracy, it’s the backbone of every successful ML system.
Without it, you’re just guessing with data.

[Jorge Valverde-Rebaza @jvalverr]

Automate pipelines with CI/CD.
Manual processes break reproducibility faster than bad data.
If your model can’t be rebuilt with one command, you’re not ready for scale.

[Jorge Valverde-Rebaza @jvalverr]

In the design and deployment of real-time systems, there is always a tradeoff: latency, privacy, and scalability. You can’t maximize one dimension without sacrificing the others; the key is to strike a balance that fits the system’s specific requirements (edge, cloud, or hybrid).

[Jorge Valverde-Rebaza @jvalverr]

If GPT-5 can remember, reason, and act proactively —

📌 How do you integrate it into operations without losing governance, traceability & risk control?

#GPT5 #GenerativeAI #PromptEngineering #AIstrategy #BusinessTransformation

[Jorge Valverde-Rebaza @jvalverr]

🚨 GPT-5 is here. And it’s not playing in the same league.
What I saw today changes how we communicate with AI — forever. 🧵

[Jorge Valverde-Rebaza @jvalverr]

The key difference:
GPT-4 → One-off shot.
GPT-5 → A living decision-making system that grows with you and your business

The new advantage is not using AI for isolated tasks…
It’s training your AI to evolve with your strategic challenges.

[Jorge Valverde-Rebaza @jvalverr]

🚀 Novidade na Newsletter! 📰 Aprenda sobre como as LLMs geram texto de alta qualidade: https://t.co/bHpYvDxamp

[Dmitry Kobak @hippopedoid]

Updated our preprint to add a map of retracted papers (11k) in PubMed (21m). There are clear clusters and we believe it's paper mill activity.

E.g. there is a small island with 11% (!) retractions; we argue that the other 89% are all VERY suspicious and deserve further scrutiny.

[Jorge Valverde-Rebaza @jvalverr]

@msobrevillac Fuerzas Marco!

[Jorge Valverde-Rebaza @jvalverr]

@Mike_Scully_ Chat

[Santiago @svpino]

I have 5 scholarships to the course "Applied Machine Learning".

Andrew Maas, Senior Manager at Apple and Instructor at Stanford University, teaches the course that starts in September 5th and lasts 4 weeks.

Read on for details on how to apply:

1 of 3

[Jorge Valverde-Rebaza @jvalverr]

@beenet_ia Ótima oportunidade pra quem tá na procura de #empregoTech com foco em #desenvolvimentoMobile, candidate-se grátis na @beenet_ia!!!