01 · Foundation
Indexed Corpus
Your manuals, SOPs, and institutional data — indexed, sovereign, and searchable as the single source of truth.
The Institutional Brain
Your organization's secure, trusted, compounding memory.
Innaris turns your indexed corpus and everyday work into a living knowledge base — every answer gated and verified against sources, every job captured as experience, so lessons learned are never repeated and patterns surface on their own.
Field technician · CH-7 chillerVerified in 6.2 s
"Low suction alarm returns after charging — what should I check?"
Grounded Verified ✓ · 89% confidence
Before adding charge again, leak‑check the evaporator header and TXV inlet — a validated field note for CH‑7 traces the recurring alarm to a header flare that was re‑flared, then recharged to 11 °F subcooling. Confirm the transducer against a gauge first; target 10–12 °F subcooling at design load.
1PROC‑2201 · Chiller low pressure procedure
2Field experience · J. Ortiz (journeyman) · validated by admin
3CH‑7 manual · nameplate figure (OCR)
Verdict: grounded · NLI passed · 5/5 claims entailed by cited sources
02 · Trust
Gated & Verified
Every response is grounded in cited sources and passes a verification gate — no ungrounded claims reach the field.
03 · Compounding
Experience Capture
Interactions become experiences — what worked, what didn't, caveats, and senior tacit knowledge, referenced on the next job.
04 · Intelligence
Pattern Discovery
Dashboards reveal common problems, solutions, and emerging themes — signalling which documents, SOPs, and training to update.
Index→Verify→Capture→Compound→Reveal Patterns
Innaris · The Authority of Memory
How it works
Retrieval that has to prove itself. Answers that carry their evidence.
A knowledge graph of your own corpus, a retrieval pipeline that must demonstrate relevance, and a delivery gate that verifies every answer before it reaches the screen.
Index everything
PDF manuals, SOPs, work orders, spreadsheets, photos. Figures inside documents are OCR'd and stored with their caption and surrounding text, so nameplates and diagrams are searchable and citable.
Retrieve with proof
Filename similarity, exact keyword grep, embedding similarity and knowledge‑graph relations — a passage only enters an answer when its own content meets the question's threshold.
Generate, then verify
The model drafts from the retrieved evidence only. An NLI verifier checks each claim against the cited passages; the answer is delivered, marked inferred, or withheld.
Capture experience
Technicians file what actually fixed it. Submissions are signed on the device, verified by an admin, and become citable memory with who, seniority, when and validation attached.
Trust
Built for liability, not just convenience.
Every answer carries the evidence behind it and an honest number for how much to trust it.
Confidence you can audit
Verified ✓ · NN% confidence
The percentage is computed from the verifier verdict, per‑claim entailment coverage, retrieval tier, evidence count and validated experience — and the tooltip lists every factor.
Always labelled
Grounded vs. inferred
No evidence in your corpus? The answer is delivered blurred as general knowledge, capped at low confidence, and revealed only on your say‑so. Alert‑linked diagnoses stay withheld until acknowledged.
Tamper‑evident
Signed and chained
Experience submissions are Ed25519‑signed and chained; the store carries a Merkle manifest. The hub refuses submissions whose signature doesn't verify.
Sovereign
Your corpus stays yours
Documents live on the device or your own Central Hub. Online mode sends only the question plus the sanitized retrieved context to the model — never your library.
Compounding
One master memory per organization.
Every install is a node. Nodes work offline and sync to your Central Hub when online, so every technician's verified experience keeps building the same institutional brain.
Verified before it counts
Local until approved
A technician's experience stays on their device and in the local review queue. It's pushed to the hub inbox as pending and joins the master graph only after an admin verifies it.
Everyone benefits
Everyone gets everyone's fixes
Nodes pull the master corpus, figures and every verified experience — with provenance — into their local store, even after weeks offline.
Governed
Users, devices, keys
Admins issue a client key per user and device, revoke it any time, and see last‑seen and sync state for the whole fleet.
Online when you want it
Cloud models on your account
The same key authorises online model calls through the hub — Amazon Bedrock, Anthropic, OpenAI and more — metered per organization with a spend cap. The picker shows exactly the models your account can use.
Plans & pricing
Three levels of sovereignty. Same Innaris on every device.
Choose Personal (a few nodes for one person) or Enterprise (many nodes across your teams) at any tier. Pricing is per active node plus the hub, scaled to the tier's capability — request access and we'll confirm a quote with your keys.
Tier 1 · Available now
Cloud Hub
Personal or Enterprise · pricing on request
Your Central Hub, master corpus and knowledge graph hosted on Innaris cloud infrastructure, isolated per organization. Cloud models (Amazon Bedrock and more) through your hub, metered to your account.
- Online data corpus & master graph, verified experiences shared across your nodes
- Cloud models with a spend cap; local model + local corpus fallback on devices with the room
- Admin approval queue, per-user keys, fleet view, audit trail
- Install on macOS, Windows, Linux, iOS, Android
Tier 2 · Coming
Protected
Enterprise · pricing on request
Everything in Cloud Hub, on infrastructure you own: a company-controlled Central Hub and data store (your premises or VPC), with an optional air-gapped server running large models behind a private, protected API.
- Company-owned hub server and store; nothing leaves your network unless you allow it
- Private large-model API on your air-gapped server (online reachability optional)
- Local device model/storage fallback when the hub is unreachable
Tier 3 · Coming
Sovereign Mesh
Enterprise · pricing on request
Everything in Protected, fully offline. Nodes run locally and the Central Hub can be offline too; nodes sync with each other over near-field / proximity links and relay changes hop by hop until they reach the hub.
- Node-to-node sync by proximity — no internet, no cellular
- Store-and-forward: changes travel across nodes to the hub and back
- For remote sites, vessels, secure facilities, disaster response
Personal = one person, a handful of devices. Enterprise = many users and nodes with roles, approval workflow and fleet management. Both start with a request below; an Innaris administrator approves it, creates your organization on the hub, and sends a one-time secure link with your admin key.
Get Innaris
Request access.
Tell us who you are and what you need. When an Innaris administrator approves the request you'll receive a one-time secure link that reveals your organization's admin key and setup steps. From there you install the app, connect it to your Central Hub, and issue keys to your own team or devices.
Download
Get Innaris on every device you use.
Desktop installs run the local model on your hardware; phones and tablets install the web app and talk to your node or Central Hub.
desktop
macOS
macOS 13+, Apple Silicon or Intel, Python 3.11+ (installer bundles it)
Runs the local model on Metal. Install as an app from Safari (File → Add to Dock) or Chrome (install icon).
From source:
git clone the repository, then `bash start_mac.sh` (creates the venv, installs requirements, starts on http://localhost:3131)
desktop
Windows
Windows 10/11 64-bit, Python 3.11+ (installer bundles it)
CUDA offload when an NVIDIA GPU is present; CPU otherwise. Install as an app from Edge or Chrome (install icon).
From source:
git clone the repository, then `powershell -ExecutionPolicy Bypass -File start_windows.ps1`
desktop
Linux
Ubuntu 22.04+/Debian/Fedora, Python 3.11+
CUDA/ROCm or CPU. Install as an app from Chrome/Chromium (install icon).
From source:
git clone the repository, then `bash start_linux.sh`
mobile
iPhone & iPad
iOS 16.4+ / iPadOS, Safari
Open your Innaris node or Central Hub address in Safari, tap Share → Add to Home Screen. Runs full-screen with offline shell; answers come from your node/hub.
mobile
Android
Android 10+, Chrome
Open your Innaris address in Chrome and tap Install (or ⋮ → Install app). Works as a standalone app.
web
ChromeOS & other browsers
Any modern browser
Use the address-bar install icon to pin Innaris as an app.
Desktop installers are published with the public release. Request access above to be first in line; your key and setup link work with the installers the day they ship.
Enterprise: your admin gives you the Innaris address (your Central Hub or a node) and a client key. Enter them in Settings → Central Hub, or in
.env as INNARIS_HUB_URL and INNARIS_HUB_CLIENT_KEY.