DataMust Labs
Experiments at the edge of practical data and AI
Field notes from what we are building, testing and learning, including what works, what breaks and what still needs evidence.
Labs documents active experiments. These field notes are observations from the work, not production recommendations or performance benchmarks.
Latest experiments
What we are testing now
In progress
Agentic AICan an AI Chief of Staff Build Its Own Team?
Inside our experiment with an AI Chief of Staff that designed and delegated work to specialist agents, including what broke and what we will test next.
Experiment
Agentic organisation design and autonomous task delegation
Hermes Agent
Multica
Multiple LLMs
Exploring a practical use case?
Discuss a governed AI use case
A 30-minute fit call helps determine whether the use case is ready to implement or needs a clearer roadmap first.