Research

How we measure agent capability: evaluation environments, benchmarks, and what they reveal.

Latest in Research

1 post

A pixel-art street scene split down the middle. On the left a warm, ordinary high street of shops; on the right the same street redrawn as a blue blueprint, its shopfronts renamed as software files. Fo stands on the dividing line between them.

The Simulated World Where Agents Act

Wajo builds Action Agents — agents that guarantee outcomes, not just tokens. Agents acting in the physical world can't be graded on text alone. So we built a world where they take real action: real accounts, synthetic websites, simulated voice callers on a foreign voice stack. Fo, one of our products, hillclimbs in it.

August 13, 2026Jeroen Olieslagers, Shivani Poddar