BridaBrida Journal
Build AI systems you can measure.
Product releases, benchmarks, engineering notes and research from Brida.
Product
Introducing Brida Reflex: fast typed decisions for agents and software
Brida Reflex turns recurring events into small, bounded decisions. Today we are releasing the public product surface together with ReflexBench, open schemas, the SDK and a growing library of use cases.
Research
Introducing ReflexBench v1: an open benchmark for System One models
Brida is open-sourcing ReflexBench v1, a reproducible benchmark for typed decision engines with calibration, multilingual, robustness and workflow-policy evaluation.
Research
Reflex Alignment and AlignmentBench: measuring AI at the moment it acts
A practical execution-control pattern and an open benchmark for testing whether models and agents stay inside explicit authority, policy, uncertainty and oversight boundaries.
Research
Auditing the training loop: a practical response to We Must Pace the Frontier
Dario Amodei argues that independent evaluators should inspect not only finished frontier models but the training pipelines that create them. Here is one narrow engineering primitive for making that idea more measurable.