Stack · v0.4 · 6 Layers
/ 06The Stack

The Eru Stack.

Six layers, end to end. From a worker's first-person capture in the field, through structured task graphs and outcome links, to deployed applications across robotics, insurance, copilots, and autonomous work systems.

/ 06.AArchitecture

Capture → Application. One model in the middle.

Every signal flows up. Every prediction flows down. Layer 5 — the Eru Physical World Model — is the gravitational center.

Eru Stack v0.4
6 Layers · 72 Modules
L1Capture

Capture

Raw signal from the physical world. First-person POV, third-person context, site sensors, partner streams.

Aegis · POVBeacon · sitesite camerasmobile uploadspartner sensorsrobotics dataIMU · forceaudio · ambient
L2Ingestion

Ingestion

Signal becomes substrate. Synchronized, filtered, permissioned, ready for structure.

video processingframe segmentationevent segmentationsensor sync ±8msmetadata alignprivacy filtersdata permissionsprovenance chain
L3Structuring

Structuring

Pixels become a graph. Tasks, tools, objects, zones, sequences — all linked into an event graph.

task detectionsequence mappingtool / object recorole attributionzone mappingevent graph buildontology align
L4Outcome Linking

Outcome Linking

The structured graph is bound to business truth. Inspections, rework, claims — Eru's source of causal signal.

inspection resultschange ordersrework eventsclaims datasafety incidentsschedule milestonescertificationsquality scores
L5Eru · Model

Physical World Model

The center of gravity. A 2.4B-parameter foundation model with perception, workflow, outcome, action, memory, and prediction heads.

perceptionworkflowoutcomeactionmemory · ∞predictioncausal headretrieval
L6Applications

Applications

Surfaces partners and customers actually use. PIN OS, insurance, robotics APIs, copilots, simulation, autonomous work.

PIN OSPIN Certifiedinsurance intelrobotics APIstraining systemsindustrial copilotssimulation engineautonomous work
/ 06.BFlows

Three flows through the stack.

The stack carries three distinct kinds of traffic, on the same six layers. Each flow has its own latency, retention, and access policy.

Flow A · Signal Up

Worker captures → events structured → outcomes linked → model updates. Latency: minutes to days. Permanence: indefinite, permissioned, revocable.

Flow B · Prediction Down

Model heads → APIs → PIN OS dispatch, copilots, robotics partners, insurer dashboards. Latency: 12–80ms. Permanence: query-time only.

Flow C · Certification Sideways

Outcome layer → PIN Certified → portable records that follow workers, machines, companies, and assets. Verifiable, signed, audited.

/ 06.CApplications

Eight surfaces built on Eru.

Layer 6 is where the model meets the world. Some are built by PIN. Some are built by partners. All draw from the same model.

App · 01
PIN OS

Operational system for physical work. Scheduling, dispatch, supervision, inspection.

App · 02
PIN Certified

Portable proof-of-work records for people, machines, and assets.

App · 03
Insurance Intelligence

Verified worksite signal feeding underwriting, prevention, and claims defense.

App · 04
Robotics APIs

Task priors, action knowledge, evaluation, and handoff for partner platforms.

App · 05
Training Systems

Human upskilling driven by expert demonstration and supervisor correction.

App · 06
Industrial Copilots

In-line guidance for trades, foremen, and shop-floor operators.

App · 07
Simulation Engines

Long-tail scenario generation mined from real work for training and stress testing.

App · 08
Autonomous Work Systems

Mixed human-robot workflows orchestrated against verified outcomes.

Pick where you plug in.

Integrations exist at L1 (sensors), L4 (operational data), L5 (model API), and L6 (applications).