Built in strict order: Ellie first, because every other agent can only exist on the foundation she provides. Ellie is complete; Mia is running on her output; the Investigator agent and industry rollout come next.
Ellie is built and measured: she reads legacy fault-tree drawings, rebuilds the Boolean structure across pages, runs structural checks and exports executable trees, at 99.96% node-level accuracy on held-out production pages. She runs today on partner archives inside air-gapped perimeters.
Mia runs on Ellie's output today: exact cut sets, importance measures, latent-fault classes, review ranking and a grounded assistant on a local model. Work continues on probabilistic quantification from un-sanitised drawings and on the fleet-wide event library.
Tracing incidents backward through Ellie's compiled failure chains and Mia's cut sets to isolate root cause — turning the same executable models into a forensic tool for fleets in service.
Ellie and Mia pilots with airframer and supplier partners on their own premises, the Investigator following on the same foundation.
We'll show you where each agent is today and what ships next.