Baru fuses every source into one trusted picture — and leaves the decision to a human. It brings sensor, signal, and contextual data together at the edge, in disconnected or contested environments, and turns them into a single operational picture with evidence behind every element. Fusion runs where the mission runs: on sovereign infrastructure, under zero-trust security, connected or not.
Baru never authorises action. It proposes, correlates, and explains; a cleared human operator decides. That line never moves.
This line holds in every Enkidu sector — see Trust.
Baru serves its sectors through named implementations — Enlil for Defence, Lamassu for urban and public safety — each tuned to its domain’s models, policies, and compliance posture on the same platform.
Long description
Four stages left to right, all recorded on one continuous chain of custody that is signed, hash-linked and tamper-evident. Ingestion is a four-step pipeline — authenticate, tag and classify, correlate, record custody — and nothing moves untagged, unauthenticated or unrecorded, under any load. Observation is the common tagged object carrying identity, origin, classification and custody. Track is correlated observations of one entity. Assessment is inference over tracks with lineage pointers back to every parent. Each stage drops a tick onto the custody band. The ladder ends at a separate block, set apart by a gap and its own border: a human decision, outside the automated flow.
Built on Ekur
Baru runs on Ekur, the shared technology stack beneath everything we build. Its capabilities are not products — they are the common foundation that makes every Enkidu platform secure, auditable, and sovereign by construction.
Sector implementations
Baru runs one specialist per modality — vision, RF, acoustic, LiDAR, text — each an open-weight base adapted, quantised and distilled for its job and its tier. Enkidu does not pre-train foundation models; it makes strong open bases excellent at your task, and commits to keeping fielded models current: monitored for drift, retrained against mined failures, re-evaluated under the zero-regression rule, and redistributed under Nusku. Which model runs on which node under which conditions is the operator’s declaration.
Enkidu is evaluating three technologies for Baru’s edge fabric. NVIDIA Halos — a functional-safety architecture that places a deterministic, hardware-isolated authority between AI-derived conclusions and consequential action. Quantization-aware distillation — a compression technique that teaches a low-precision model to remain faithful to its full-precision parent, fitting reasoning models to the edge. Speculative decoding — an inference-acceleration technique for reasoning models at the edge.
Evaluation and prototyping activity; no availability or performance claims.