Perpetua
A managed, persistent private database brought to a site over a single encrypted port — with a localized-intelligence database manager that keeps it organized, indexed, and healthy. The durable data layer that Intelligence on Demand answers from.
Data that stays, and stays organized
Every private-AI capability needs somewhere durable to keep what it knows. Perpetua is that place. It brings a managed database to a site and keeps it — persistent, indexed, and consistent — so the corpus a model works from does not live in a spreadsheet, a scattered share, or someone's laptop. The name says the job: data that endures.
What makes Perpetua different from "just a database" is the manager that comes with it. A localized advanced intelligence runs alongside the data, on the same hardware, and does the ongoing work a database usually waits for a person to do: organizing new material as it arrives, keeping the index current, reconciling duplicates, and watching for the drift and decay that quietly rot a data store over time.
How it works
One port in, a persistent database, and a manager that never goes home. Everything lives inside the boundary you control.
One encrypted port
Perpetua reaches a site over a single encrypted port — one way in, identity-controlled and auditable. Simple to allow, simple to reason about, with no sprawl of exposed services to secure.
A persistent database
A real, managed database stands up at the far end and stays. Your corpus is held durably and consistently inside the boundary, not sent out to a service that holds it for you.
A localized-intelligence manager
An on-site intelligence keeps the store organized: it ingests and classifies new material, maintains the index, reconciles duplicates, and flags drift — the housekeeping a database usually waits on a human to do.
The layer beneath Intelligence on Demand
Perpetua and Intelligence on Demand are built to work as a pair. Perpetua keeps the data; Intelligence on Demand recalls it. Perpetua is always there, holding the corpus and keeping it in order. Intelligence on Demand is brought up on demand to answer a question grounded in that corpus, cite its sources, and stand down again.
One remembers, the other recalls — and neither ever sends your data outside the boundary to do its job.
You can run Perpetua on its own as a managed private database, or pair it with Intelligence on Demand to get grounded, source-cited answers over the data it keeps. Either way it is built on our private AI hosting model: the data, the index, and the intelligence stay on hardware you control.
Managed with judgment, not left to drift
Perpetua is built the way everything here is: with augmented engineering. The intelligence does the tireless, repetitive work of keeping a large store clean and current at a scale no one would do by hand; an engineer still sets what good organization means and verifies the result. The leverage is the model; the judgment stays human.
Need data that stays put and stays private?
Tell us what the corpus is and where it has to live. We will tell you where Perpetua fits — on its own, or under Intelligence on Demand.
Email info@mcgllc.com