Self-hosted AI products.Zero telemetry.
A private AI workspace, meeting transcription, and neuro-analytics for creative. Deployed in your VPC, on-prem, or fully air-gapped.
One deployment model.
One security review.
Each product installs the same way, into your VPC, on-prem, or air-gapped. Run one of them or all three on the same stack.
Lumen
Chat, RAG, and agents on your servers. 200+ models, team memory, full admin control.
Notetaker
Real-time meeting transcription and summaries on our own 4B speech model.
Noesis
Predicted cortical response for video, audio, and copy, before you ship.
You own the data.
Cloud AI does not fit.
Same work, two operating models. What changes is where the data sits, what it costs at scale, and whose controls your auditors end up reviewing.
Two weeks to a pilot.
In your own VPC.
Four steps from the first call to production. Pilot proposal in 48 hours, no commitment before the pilot, and you keep the runbook.
- 01
Discovery call
30 minutes, no slides. Integration surface, security constraints, who signs off.
- 02
Security review
Your CISO's questions answered in writing: architecture, data flow, threat model.
- 03
Pilot install
Containerized deployment to staging in your VPC. Two weeks median, one engineer.
- 04
Production rollout
SSO, SCIM, runbook handoff. We stay on call for the first 90 days, then it is yours.
The questions that come up
in every security review.
Pricing, hardware, air-gapped operation, and what happens to your deployment if we are not around.
Flat annual license per product, unlimited users on Business and Enterprise tiers. Starter is a monthly plan for teams under 20. Running both Lumen and Notetaker earns a 15 to 20% bundle discount, since they share the same infrastructure stack.
Pilot install: two weeks median, containerized into a staging environment in your VPC. Production rollout with SSO, SCIM, and integrations: 4 to 6 weeks. A named engineer stays on call for the first 90 days.
A VM running Docker Compose covers the core stack. Fully self-hosted inference for Notetaker needs a GPU; both products can route to a managed model API while your hardware is provisioned. We review and size your setup before the pilot.
Yes. Replace cloud model access with self-hosted Ollama or vLLM instances and every product operates with no external network access, zero telemetry, and no phone-home.
Enterprise: white-glove deployment, a dedicated Slack or Teams channel, and a priority SLA. Business: email and ticket support with 48-hour response. Starter: documentation and community support. All tiers get updates and security patches for the license period.
Source-available core, customer-owned deployment, optional code escrow on enterprise contracts. You keep running with no vendor dependency.
“Every team we work with hits the same wall: the AI tools they want are SaaS-only, and the data they want to use is the data they can't send to a vendor. So we built tools that run where your data already lives. That's the product.”
Sebastian Mondragon, Founder of Particula Tech