Documentation

Start here. The detail lives inside the product.

This page covers getting started, where to find the full documentation, and the questions people ask before they try Snixia Insight. The complete administrator, API and production guides ship inside the image at /docs/ in English and Simplified Chinese — available offline, with no internet access needed.

i
This page is the starting point, not the full manual. Detailed configuration, API and deployment references are delivered inside the product. If you are evaluating the business case, the product overview is the quickest read.

Get running in 30 seconds

One self-hosted install on a server you already own. No new database, no external service and no signup. A 30-day full-feature trial starts automatically.

Do not mount an empty config/ directory. The image ships its built-in configuration there. Copy those files out first, or the container will refuse to start.
terminal — 30-second start
# 1. Copy the five built-in config files out of the image
mkdir -p config data logs
for f in app.yaml connections.yaml policies.yaml chart_rules.yaml license_public_key.pem; do
  docker run --rm snixer/snixia-insight@sha256:616a02226a75929dbaa87392547ec42aa3ed0b356b9d9c46fc1037724ccea1ea \
    cat "/app/config/$f" > "config/$f"
done
sudo chown -R 1000:1000 config data logs

# 2. Start the container
docker run -d --name snixia-insight -p 8080:8080 \
  -v $(pwd)/config:/app/config \
  -v $(pwd)/data:/app/data \
  -v $(pwd)/logs:/app/logs \
  snixer/snixia-insight@sha256:616a02226a75929dbaa87392547ec42aa3ed0b356b9d9c46fc1037724ccea1ea

# 3. Open http://localhost:8080/setup

Then open the setup console in a browser and ask your first question against the bundled sample reports. When you are ready, point the configuration at your own reports.

Everything beyond this point — connection setup, report configuration, API integration, production sizing and day-to-day operations — is covered by the documentation built into the image.

Where the full documentation lives

The complete documentation is delivered inside the product at /docs/. It is available offline, in English and Simplified Chinese, and covers the end-user guide, administrator configuration, the feature catalogue, adding and configuring reports, API integration, production deployment, and the legal / data-flow / erasure policy.

The cards below are quick pointers. Detailed steps, field references and configuration examples are kept in the built-in guides rather than published here.

API & integration

Ask questions and run reports programmatically. Endpoints, payloads and authentication are documented in the built-in API guide.

Request integration details →

Adding a report

A new report is a configuration change, not a development project — no developer, no deployment window, no downtime. Templates and field descriptions are in the built-in guide.

Your own reports

Customer-owned reports can be connected, including custom ones. Step-by-step setup is in the built-in administrator documentation.

Going to production

Start with one small install and grow to multiple sites when you need to. Sizing, security hardening and the go-live checklist are in the built-in production guide.

Read the security & compliance summary →

Questions people ask before trying it

For a short overview, see the FAQ on the product page. This section goes deeper into the questions that come up while evaluating — especially around data, trust and what happens after the trial.

What it is
Is this another BI dashboard or reporting platform?

No. It is a question-and-answer layer on top of the reports you already have. It does not replace your reporting stack, and there is no new dashboard to build or maintain.

Do we need to replace or migrate our existing reports?

No. Nothing is migrated. Answers come from reports that already exist in your system, and you keep using them exactly as before.

What do we actually get on day one?

One self-hosted install, a set of sample reports to try it with, the full documentation built in, and a 30-day free trial that starts automatically. No registration and no sales call.

Data & security
Will our ERP data be sent to a vendor cloud?

No. Answers are produced inside your own environment. No copy of your business data is stored with a vendor, and nothing needs to leave your network.

Does the AI model see our data, or use it for training?

The model helps phrase and route the question; your business data is not used for training. You can use your own model or a locally hosted one — the choice is yours.

Could someone see a report they are not authorised for?

No. The access rules and approval boundaries you already rely on are applied before anything is read. Requests outside those rules are stopped, not just logged.

Does it work in a site with no internet access?

Yes. Installation, documentation, licensing and evaluation all work offline. Offline installation bundles are available for regulated or disconnected environments.

Trusting the answers
Can the AI invent numbers that are not really there?

No. It does not generate figures. An answer is built from the result of an existing report, and its source is shown alongside it — so a number can always be traced back to where it came from.

How do we know which report and which rules produced an answer?

Every answer comes with its source and a complete, tamper-proof execution record that explains how it was reached. That record is exportable for review.

What if a question cannot be answered?

It fails clearly and gives a reason. It does not guess or return a plausible-looking but wrong result.

Who can review the history of questions and answers?

Only authorised reviewers. The full record can be exported for internal audit, an external auditor or a regulator.

Setup & operations
What do we need to install, and how long does it take?

One self-hosted install, roughly 30 seconds, on a server you already own. No mandatory additional database and no external service to sign up for.

Do we need a consultant or an implementation project?

No. There is no implementation project, no custom code and nothing to transport into your ERP.

We have several systems, plants or company codes. Can it handle that?

Yes. Start with one connection and add more as needed; it can grow to multiple sites and shared infrastructure. The details are in the built-in administrator documentation.

What happens when we upgrade our ERP?

Nothing changes on your side. The product does not modify your ERP, so your upgrade cycle is untouched.

Do we have to install any additional SAP software?

No. You choose the connection method that fits your landscape — standard web services, OData or your existing report endpoints — and most deployments need nothing extra at all. One optional single sign-on path may use a component you obtain yourself; it is never bundled with the product.

Trial, licensing & support
Do we need to register or talk to sales to try it?

No. The 30-day full-feature trial starts automatically on install — no email, no license file and no activation call.

What happens after 30 days? Is our data deleted?

The system becomes readable-only: answers and exports keep working and history is retained. Nothing is erased. Activate a license whenever you are ready to restore full capacity.

What if we decide to stop using it?

Remove the container. Your data stays in your own volumes, and there is no vendor-side copy to clean up.

How do we get help and updates?

Commercial licenses include support and version upgrades. See Pricing, or contact us directly.

Still have a question? Most "how exactly does this work" answers are in the documentation built into the product, which you can read before connecting any real system. For anything commercial, get in touch.

Start with the trial. The manuals come with it.

One install, a 30-day free trial that starts automatically, and the complete documentation available offline at /docs/.

Back to quick start See it in action