Secure Environment AccessCIN-IDP

Access to this environment is governed by Companion evaluation policies and applicable confidentiality obligations. All access events are logged and monitored.

Email Address
A one-time authentication code will be issued to your registered address. Access is monitored and least-privilege scoped.
One-Time Authentication Code
issued to {{ authEmail }}
Code expires in 09:58 · Resend
Provisioning environment{{ authStep }}/{{ authTotal }}
{{ l.mark }}{{ l.t }}
Operational Context Resolved
{{ c.k }}{{ c.s }}{{ c.v }}
AUTHORIZED OPERATIONAL USE ONLY · Access may be monitored, rate-limited, version-controlled, or revoked at Companion Health's discretion. Subject to platform governance and applicable agreements.
Fetch API Canonical Intelligence Exchange
Companion HealthCompanion Health
The thesis

No single organisation sees the whole pet.

Health is continuous. Observation is fragmented. Each system holds a true but partial view — and none of them, alone, can see the animal.

{{ d.sys }} see only {{ d.sees }}
Companion exists to

reconstruct continuity across fragmented observation systems.

One canonical thread per animal — resolved, qualified and made interrogable — so the signals that were always there, scattered across systems, finally become visible in time to matter.

Animals cannot describe what they feel. Continuity is the closest thing they have to a voice — and the reason this system exists.

Scale & Governance Center

Can this run?

Live operational posture across the network. Figures are representative of production-scale operation.

{{ m.k }}
{{ m.v }}
{{ m.s }}
Encryption · in transit + at rest Audit · immutable · 100% coverage Multi-tenant isolation · per environment Partner-retained data ownership
CIN Membership Center

How to join the network.

A defined path from qualification to optimization. Partners retain ownership of source data throughout; value compounds with continuity.

{{ c.n }}{{ c.name }}{{ c.dur }}

{{ c.detail }}

Indicative durations. Onboarding is governed by the CIN participation agreement, DPA and applicable data-sharing terms. Source data remains owned and controlled by the contributing environment.
Infrastructure Roadmap

What is built, and what comes next.

An infrastructure roadmap, not a marketing one. Continuity compounds before intelligence builds on it.

Current · live
{{ r.name }}{{ r.status }}
Next · sequenced
{{ r.name }}{{ r.status }}
FORWARD-LOOKING — Sequence and timing reflect current development intent and are subject to change. Certain layers represent target-state capabilities not yet generally available.
FetchCanonical Intelligence Exchange
Companion Health Establishing operational context {{ integEnvName }} {{ integPct }}
Observational records ingested
{{ integTotal }}
{{ integStageLabel }} · resolving into one canonical substrate
{{ s.mark }}{{ s.label }}
Active IntegrationsFetch endpoint registration
{{ g.brandEl }}{{ g.kind }} {{ g.recs }}{{ g.status }}
Records are identity-resolved into canonical subjects, then qualified for signal formation. Source systems retain ownership.
AUTHORIZED OPERATIONAL USE ONLY · Integration activation reflects representative operational scenarios. Signals, interpretations and continuity artifacts are derived from available observations and may evolve as additional information becomes available.
Descend · canonical object
{{ drawerSid }}

{{ drawerCls }}

{{ drawerName }} · {{ drawerBreed }}
STATUS
{{ drawerStatus }}
SCORE
{{ drawerScore }} · {{ drawerBand }}
CONFIDENCE
{{ drawerConf }}
PERSISTENCE
{{ drawerPersist }}
Derived from · {{ drawerEvCount }} contributing observations
{{ e.day }} {{ e.env }} {{ e.kind }} {{ e.state }}
{{ e.w }} {{ e.conf }}
Conflict reconciliation
{{ c.what }}{{ c.state }}
{{ c.at }} · {{ c.res }}
Source environments
{{ s }}
Who consumed this signal
One canonical object · {{ drawerEvCount }} interceptions · 0 duplications. Each consumer reads the same signal_id through a different lens.
What value this signal created
{{ v.k }}
{{ v.v }}
{{ v.s }}
PROVENANCE — Every attribute traces to source, time, confidence and transformation. Signals are derived from available observations and may evolve as new information arrives.
Interrogate · {{ conceptKind }}
{{ conceptTitle }}
What is this?

{{ conceptSummary }}

Continue to Explanation for why it exists and how it works, then Evidence for the data behind it.
Why it exists · how it works

{{ p }}

Evidence · {{ conceptEvCount }} records
{{ e.k }}{{ e.v }}
Evidence is not the end of the chain. Continue to Source for the actual object.
Source · {{ conceptSrcKind }}{{ conceptSrcId }}
{{ conceptSrcId }}
{{ s.k }}{{ s.v }}
Terminal object. This is the actual record the assertion resolves to — not a summary of it. Source chain complete.
RETRIEVING OBJECT…
FetchCanonical Intelligence Exchange
Intelligence service
{{ iStageVM.name }}AI service
{{ iStageVM.svc }}
{{ r.label }}

{{ r.body }}

Every stage is explainable and reconstructible. Evidence terminates in records traceable to partner source.
FetchCompanion Intelligence Layer
Active process
{{ inspectJob.name }}
{{ inspectJob.svc }}
Current activity
{{ r.k }}
{{ r.v }}
Inspection · live reconstructionconf {{ inspectConf }}
{{ e.kind }} {{ e.t }}{{ e.meta }} {{ e.conf }}
RECONSTRUCTED — Not a linear pipeline: observations competed, a conflict was reconciled by a late arrival, and confidence was adjusted as evidence accumulated. This process runs continuously across every active subject.
FetchContinuously reconstructing continuity
Source Trace · {{ traceKind }}
{{ traceTitle }}
Trace chain · resolved to source
{{ c.layer }}
{{ it.k }}{{ it.v }}
Chain complete — assertion resolves to partner source. Nothing here is generated; everything is derived from a held observation with provenance.
Retrieving · resolving source
{{ r.mark }}{{ r.t }}
FetchProvenance · source-resolved
← Companion PlatformDownload PDF ↓
Companion Health
LENS {{ lensName }}
COHORT{{ tCohort }}
SIGNALS{{ tSignals }}
RISK>60{{ tRisk }}
CONF{{ tConf }}
LAG{{ tLag }}
MODELckd-v1.0
{{ clock }}UTC · SYNC {{ sync }}
LIVE
MULTI-TENANT CONTEXTsubstrate persists ·
SUBSTRATE
{{ residue }} {{ ingRate }}
Computing
{{ b.mark }}{{ b.t }}
{{ envName }} · {{ envMode }} · runtime live00 / RUNTIME

The ecosystem is already operational.

Continuity is accumulating across ingestion surfaces, partner environments and behavioural domains — independent of inspection. Companion does not generate signals. It resolves them.

{{ m.k }}
{{ m.v }}
{{ m.s }}
COHORT · live subjects{{ cohortMeta }}
INTERPRETIVE LAYER

The runtime persists whether inspected or not. Open the workbench to trace how one signal forms, qualifies, and drives downstream impact.

INGEST · packet lanes{{ ingRate }}
● telemetry● clinical● claims● intake
{{ ingestNote }}
Companion Intelligence Network01 / SUBSTRATE

One substrate, beneath the surfaces.

Companion is the interpretation layer the ecosystem integrates — not a destination. Each environment contributes a different observational geometry; none sees the whole pet alone. The canonical substrate persists across all of them.

{{ f.n }}
{{ f.label }}
{{ f.meta }}
ACTIVE ENVIRONMENTS · production state
Continuity · longitudinal thread02 / SUBJECT

{{ subjName }}. One persistent thread.

{{ subjMeta }}

The same identity, records, behaviour and history each surface already holds become one continuous thread, followed through every layer below. Observations arrive late, incomplete and sometimes in conflict — the lattice holds each with its provenance.

Subject Continuity Record · why {{ subjName }}{{ subjRecordId }}
Identity sources · connected systems
{{ s }}
Confidence trajectory{{ subjRecordConf }}
{{ c.k }}
{{ c.v }}
OBSERVATION LATTICE · 90dprovenance held
{{ o.day }} {{ o.dom }} {{ o.label }}{{ o.flag }} {{ o.state }} {{ o.conf }}
OBSERVATIONAL DOMAINS
{{ d.name }}
{{ d.kind }} · {{ d.cadence }}
The Signal Workbench · trace · qualify · prove03 / WORKBENCH

Trace a signal to its origin.

Not a card — an environment. Follow how a signal forms from raw observation, the criteria that qualify it, the impact it triggers, and who consumes it. This is where interpretation becomes accountable.

Signal genesis · 90-day formation lineage
{{ g.day }} {{ g.env }} {{ g.reading }} {{ g.verdict }} {{ g.conf }}
The signal did not appear — it accumulated. No canonical object exists until confidence crosses 0.70 with sustained persistence.
Intelligence ContributionAI services applied to this signal
{{ a }}
Each service is explainable and confidence-disclosed. Open the Intelligence Layer to inspect any service.
Qualification criteria · why this resolved to a signal
{{ q.mark }}{{ q.k }}{{ q.v }}
{{ q.formula }}
{{ q.detail }}
ALL CRITERIA METQualification is mathematical, not declarative — every threshold above is computed from the evidence lineage and recomputed as new observations arrive.
Signal impact · what happened because it emerged
{{ m.what }}{{ m.detail }} {{ m.metric }}
Outcome chain · signal → interpretation → action → outcome
{{ o.stage }}
{{ o.label }}
Signal consumers · who read this canonical object
One canonical object · multiple consumers · zero duplications. Each consumer reads the same signal_id through a different interpretive lens.
The Signal Engine · canonical formation03 / SIGNAL

A signal emerges from the field.

Persistence qualifies. Velocity sharpens. Confidence crosses threshold. Only now does a named signal resolve — derived from the substrate, never generated by a core. Risk appears here for the first time, and only here.

FORMATION FIELD · confidence over 90d{{ sigStateLabel }}
{{ b.day }}
— — — threshold · confidence 0.70 — — —
{{ g.label }}
SIGNAL OBJECT · click to descend ›SIG-204-RENAL
RESOLVED CLASS
Renal trajectory
SCORE
{{ sigScore }}
CONFIDENCE
{{ sigConf }}
{{ sigStateLabel }}
LineageQualified
Subject continuityEstablished
Interpretation stateActive
PROVENANCE · contributing
{{ p.name }}{{ p.w }}
Evidentiary lineage · meaning reconstruction04 / REPLAY

Rewind the belief.

Replay reconstructs how the conclusion formed — which observations mattered, when confidence crossed, how a late vet weight resolved an earlier conflict. Not a log. An evidentiary record.

T−86dconflict −19dresolve −14dNOW
REPLAY HEAD
{{ replayDayLabel }}
RECONSTRUCTED EVIDENCE · click to inspect{{ replayObsCount }} observations
EVIDENCE INSPECTOR{{ insDay }}
{{ insLabel }}
Source
{{ insSource }}
Provider
{{ insProvider }}
Mode
{{ insMode }}
Contribution
{{ insWeight }}
State
{{ insState }}
Confidence
{{ insConf }}
BELIEF STATE · at head
SCORE
{{ beliefScore }}
CONF
{{ beliefConf }}
{{ r.k }}{{ r.v }}
OBSERVED
{{ observedCount }}
INFERRED
{{ inferredCount }}
One substrate · many truths05 / INTERPRETATION

The same signal, read five ways.

The Interpretation Layer composes consumer-specific Briefs from identical primitives. One signal_id, never duplicated. The substrate does not change — only the lens changes.

CANONICAL OBJECTimmutable
SIG-204-RENAL
Renal trajectory
SCORE
58
CONF
0.84
INTERCEPTIONS · 5 · 0 DUPLICATIONS
{{ x.at }}{{ x.surface }}{{ x.conf }}
{{ briefWho }} BRIEF{{ briefSurface }}
{{ briefHeadline }}

{{ briefBody }}

{{ f.k }}
{{ f.v }}
Interpretation construction pipeline
{{ p.t }}
How this brief was constructed{{ briefWho }} lens
Signals used · reweighted
{{ s.k }}{{ s.w }}
Evidence drawn
{{ e }}
Time horizon
{{ constructHorizon }}
Confidence basis
{{ constructConfidence }}
Rationale · {{ constructRationale }}
SUBSTRATE UNCHANGEDEvery brief above derives from the one canonical object on the left. Nothing was copied; only the lens was reweighted.
INTERPRETATION LAYER — Interpretations represent contextualized views of canonical signals within a specific operational lens. Interpretations may vary across environments while remaining linked to the same underlying signal object.
Delivered throughFetch API06 / FETCH

What the endpoint consumes.

Five endpoints, one contract. The environment streams signals and reads interpreted state — payloads carrying confidence, timestamps, lineage and inferred-vs-observed markers. No platform replacement, no workflow disruption.

Canonical intelligence distributionLIVE PROCESS
{{ d.label }}
{{ l.iconEl }}
Fetch API
Canonical Intelligence Exchange

Partner systems emit observations. The substrate resolves identity, stores canonical objects and qualifies signals. Fetch API is the single contract through which every consumer reads interpreted state — upward through the stack, never around it.

observations rise
signals qualify
intelligence delivered
Fetch API Gateway telemetry · live OPERATIONAL
{{ s.k }}
{{ s.v }}
{{ s.s }}
Operations center · live request funnel
{{ f.k }}{{ f.v }}
Records resolve to identity, deduplicate, and condense into canonical objects. Only a fraction become signals; each fans out to many interpreted, delivered responses.
{{ a.k }}
{{ a.v }}
PUSH + PULLStreaming subscriptions and request/response over one versioned, multi-tenant contract.
Request lifecycle · single call
GET /v1/health-summarytrace · {{ tLag }}
{{ r.step }} {{ r.detail }} {{ r.ms }}
Endpoints · canonical contract
TO INTEGRATE
4 endpoints
TO FIRST SIGNAL
2 weeks
{{ payloadMethod }} {{ payloadPath }}200 · {{ tLag }}
{{ payload }}
Fetch API
Canonical Intelligence Exchange
© 2026 Companion Signal Intelligence Labs LLP
Developer Notice

Use of Fetch API is governed by Companion licensing terms and applicable data-sharing agreements. Returned intelligence objects may contain derived, inferred, reconciled, or confidence-qualified attributes that should be evaluated within the context of the consuming environment.

FETCH API DISCLAIMER — Fetch API delivers canonical intelligence objects, continuity artifacts, confidence metadata, provenance records, and interpretation payloads. Fetch API does not provide diagnostic conclusions, treatment recommendations, underwriting decisions, or automated determinations. Consumers remain responsible for independent evaluation and use of delivered information. DATA RIGHTS & ATTRIBUTION — Companion preserves attribution to originating observations wherever technically feasible. Derived signals, confidence measures, continuity artifacts, and intelligence objects may incorporate observations contributed by multiple participating environments. Rights associated with underlying observations remain subject to applicable agreements and governing policies.
Applied intelligence · why the substrate resolves07 / RESOLUTION

Continuity reveals emergence.

Companion does not predict disease — it reconstructs continuity, and continuity reveals emergence. Select an audience, then interrogate any capability to its source.

{{ ucActiveLabel }} · resolved capabilities · click to interrogate
Each capability draws on the same canonical substrate — no separate model per use case. Click any capability to trace why Companion reached it, from interpretation to partner source.
Capability Resolution Record
{{ capName }}
Resolving continuity record
{{ r.mark }}{{ r.t }}…
Signal Genesis
{{ capSigName }}
First observed{{ capSigFirst }}
Persistence{{ capSigPersist }}
Threshold exceeded{{ capSigExceed }}
Canonical confidence{{ capSigConf }}
{{ s.k }}
{{ s.v }}
Contributing signals · why Companion reached this
Click any signal for its genesis, then trace to partner source. Continuity → interpretation → signal → derived event → raw observation → partner source.
FetchResolved on canonical subject {{ capSubjectId }}
Continuity compounding · why the network gets smarter07 / COMPOUNDING

Why the network gets smarter.

The moat is not AI, signals or interpretations. The moat is time. Move the timeline — watch continuity, confidence and intelligence compound as observation density accumulates.

{{ cmpLabel }} network age · {{ cmpDay }}
{{ m.k }}
{{ m.v }}
What compounds with time
{{ b.k }}{{ b.v }}
The advantage a competitor cannot buy
New entrant · cold start
Subjects{{ cmpColdSubjects }}
Signals{{ cmpColdSignals }}
Confidence{{ cmpColdConf }}
vs
Companion · at {{ cmpDay }}
Subjects{{ cmpNowSubjects }}
Signals{{ cmpNowSignals }}
Confidence{{ cmpNowConf }}
The input is time, which cannot be purchased. A competitor with more capital still cannot retroactively acquire the years of resolved, canonical continuity the network compounds. Figures are representative of network-scale operation.
Engineering release history · platform construction09 / RELEASES

What it took to build this.

Not a marketing roadmap — a release ledger. Each shipped release opens its build record: duration, hours, what was constructed, what failed first, and the moat it created.

{{ m.v }}
{{ m.k }}
Release ledger · click a shipped release to open its build record
FORWARD-LOOKING — Shipped releases reflect delivered platform capability. Planned layers represent current development intent and target-state capabilities; sequence and timing are subject to change.
Build record · {{ relV }} · {{ relDate }}
{{ relName }}

{{ relSummary }}

{{ s.k }}
{{ s.v }}
Approaches · what failed first
{{ a.label }}{{ a.outcome }}
What was built
{{ b }}
Before
{{ x }}
After
{{ x }}
Evidence
{{ e.k }}{{ e.v }}
Strategic value · the moat

{{ relValue }}

Trust, Governance & Provenance · responsible intelligence08 / TRUST

Built for responsible intelligence.

Partners retain ownership of source data. Intelligence is created under privacy-by-design, hardened security, and disclosed scientific confidence. Trust scores below reflect the governance posture of this operational environment.

Legal Privacy Security Data Governance Provenance AI Principles Responsible Intelligence Compliance Contact
Assurance Index
{{ trustIdx }}
/ 100
Posture · operational
Encryption
AES-256
transit + at rest
Audit coverage
100%
immutable log
Confidence
disclosed
every signal
{{ trustLayerName }}
{{ trustLayerSummary }}
{{ c.k }}{{ c.v }}
AUDIT LOG · liveimmutable
{{ a.t }}{{ a.ev }}{{ a.detail }}
DATA RIGHTS & ATTRIBUTION — Companion preserves attribution to originating observations wherever technically feasible. Derived signals, confidence measures, continuity artifacts and intelligence objects may incorporate observations contributed by multiple participating environments. Source data remains owned and controlled by the contributing environment; rights associated with underlying observations remain subject to applicable agreements and governing policies.
Companion Intelligence Layer · how continuity is reconstructed03 / INTELLIGENCE

The intelligence beneath the platform.

Companion is AI-enabled continuity infrastructure. These services are already operating — continuously reconstructing continuity across the network. This is an operations center, not a demo: inspect any running process to see it work.

INTELLIGENCE ACTIVITY · LIVE Nine services running continuously across the substrate. Counts update in real time. inspect a process ›
Why intelligence was requiredfor this one subject
{{ w.k }}
{{ w.v }}
A rules engine cannot reconcile delayed, conflicting and missing observations across four systems, nor reconstruct a timeline or compare against a cohort. This problem could not be solved without intelligence services.
AI-ENABLED, NOT AI-LED — AI services enable normalisation, resolution, qualification, interpretation, confidence and cohort reasoning. They do not replace clinical judgement, fabricate observations, or issue diagnoses. Every output is explainable, confidence-disclosed and reconstructible to source.
Questions we have been asked · objection becomes guided tour11 / QUESTIONS

Ask, and the runtime answers itself.

Each question navigates the live system to the place it is answered — and opens the exact layer. Nothing is asserted here that cannot be demonstrated there.

Investor & partner questions
Procurement & CTO questions
Investor questions establish the thesis; procurement questions establish operability. Every answer resolves to a live layer of this runtime — the Canonical Intelligence Exchange and Signal Workbench carry the procurement burden.
Fetch Exchange · Connected Deliveries today{{ fxDeliveries }} Avg response{{ fxLatency }} Active consumers{{ fxConsumers }} Fetch API