On-Prem AI Edition (Export)
AI Store Assistant — On-Prem Edition
Runs on your own hardware. Zero data leaves your network.
For stores that cannot send customer data to an external API. The assistant, the model and the database stay inside your infrastructure — with an action log you can export for audit.
Licence options
| Edition | Price (USD) | What you get |
|---|---|---|
| Cloud / BYOK | $99 | Use your own API key. Perpetual licence + 12 months of updates. |
| On-Prem | $199 | Local model (Ollama, LM Studio, vLLM). No external API required. |
| Bundle | $249 | Both editions, switch at any time. |
Remote installation $99 · Additional language pack $39 (RTL $59) · Custom AI skill training $199 per module · Updates after year one $49/year.
Minimum hardware (you buy locally)
- Entry: mini PC, 32 GB RAM, no GPU — 7–8B model, 1–2 concurrent users (approx. $450–700)
- Recommended: RTX 4060 Ti / 5060 Ti 16 GB, 32 GB RAM, 1 TB NVMe — 14B model, 3–5 users (approx. $900–1,200)
- Pro: RTX 4090 / 5090 24–32 GB — 32B model, 8–15 users (approx. $2,000–3,000)
We license software only. We do not sell, ship or warrant hardware.
Governance built in
- SQL-first: numbers are computed in your own database; the model only translates the question into SQL.
- Action log: every query, every table read, and every byte sent out — append-only, exportable as CSV or JSON.
- One-click stop and human approval for price changes, deletions and refunds.
What we do not claim
- “AI is always safe” — safety is architecture, permissions, logging and contract combined.
- “Local is always safe” — local solves egress, not insider risk.
- “A GDPR module means compliant” — compliance needs records, a DPA and a breach process.
Request an on-prem consultation
Tell us your store size and whether data must stay on your network. We reply within one business day with a recommended specification.
Practical guidance only; not legal advice. Contract and compliance requirements depend on your jurisdiction.
