Prototype — concept, not a live service. The blog is part of the redesign proposal demonstrated in this MVP and is not a current service of LIACC. Any data shown is illustrative. See the project report for context.

19 posts

research
10 Apr 2026 · 8 min

The EU AI Act at enforcement: a lab's field guide

The European AI Act has teeth in 2026. Here's the working checklist we use before a model touches a hospital, a municipality, or a grid operator.

LIACC
applied
05 Apr 2026 · 7 min

Small language models are eating the enterprise

The frontier-model arms race dominates Twitter. The real deployments we see in Portuguese hospitals and municipalities are almost entirely 3B–14B models.

LIACC
research
28 Mar 2026 · 7 min

Reasoning models vs retrieval: when test-time compute wins

Two families of techniques — retrieval-augmented generation and reasoning models — are often pitted against each other. The truth is they solve different problems.

LIACC
applied
20 Mar 2026 · 8 min

Clinical LLMs in Portuguese hospitals: the factuality problem

Deploying a Portuguese-language LLM alongside a radiologist taught us that factuality isn't an evaluation metric — it's a product requirement.

LIACC
applied
12 Mar 2026 · 7 min

Concept drift in production: keeping maintenance models honest

Concept drift is the quiet killer of predictive-maintenance models on live industrial data. A worked example of the detection stack that catches it.

LIACC
research
03 Mar 2026 · 7 min

Safety audits for multi-agent systems: a practitioner's checklist

Multi-agent systems exhibit emergent failure modes that single-agent systems cannot. Here is the audit we run before one leaves the lab.

LIACC
research
24 Feb 2026 · 6 min

Sim-to-real at the edge: the robot-learning bottleneck no one talks about

The sim-to-real literature is full of clever domain-adaptation tricks. The bottleneck we hit in the LIACC robotics lab is simpler: the real robot is slow.

LIACC
tutorial
14 Feb 2026 · 4 min

Five LLM prompts we use weekly at LIACC

Five prompts we copy-paste around the lab. The point isn't that they are magic — it's that they give you something to iterate on.

LIACC
tutorial
05 Feb 2026 · 3 min

The research engineer's toolbox, 2026 edition

The tools we actually reach for in 2026. Nothing speculative — all shipping, all in our daily loops.

LIACC
lab-notes
28 Jan 2026 · 4 min

What I learned debugging a 70B model at 3am

The story of a bad deploy and what it taught me. The short version: logs you wrote three months ago save future-you.

LIACC
lab-notes
20 Jan 2026 · 3 min

Inside a RoboCup week — a photo diary

What a RoboCup week looks like from inside a team: a short, illustrative photo diary of a run and the year of work behind it.

LIACC
opinion
12 Jan 2026 · 4 min

If you're new to AI research: start here

If you are an MSc or first-year PhD student at Porto or elsewhere, here is the short path I'd give you for your first twelve months.

LIACC
tutorial
02 Jan 2026 · 3 min

Open-weight models we actually ship in 2026

A short, opinionated rundown of which open-weight models we deploy at LIACC in 2026, and for what.

LIACC
lab-notes
15 Dec 2025 · 3 min

A day inside the Porto Ring Road digital twin

An illustrative tour of what a Porto Ring Road digital twin looks like from the inside.

LIACC
tutorial
05 Dec 2025 · 4 min

Twelve Portuguese NLP benchmarks you should know

A curated list of twelve Portuguese NLP benchmarks. If you evaluate on Portuguese, one of these is probably on your shortlist.

LIACC
opinion
20 Nov 2025 · 4 min

AI for the rest of us: nine use cases that ship in 2026

Nine concrete AI use cases I've seen ship in small Portuguese teams over the last twelve months.

LIACC
opinion
05 Nov 2025 · 3 min

How I quit writing slides (and picked up a static site generator)

A short, opinionated essay on why I replaced Keynote with a static site generator for most of my internal presentations.

LIACC
opinion
22 Oct 2025 · 4 min

Thesis supervisors want you to know this

Six unwritten rules of writing a good master's thesis that supervisors wish they didn't have to repeat every year.

LIACC
opinion
08 Oct 2025 · 3 min

When to ignore benchmarks

Public benchmarks rank models on the wrong things for most deployments. A short guide on when to look away.

LIACC