Frequently asked questions
Company-level questions. For plans, setup and security of the AI worker, see the AI Office FAQ.
Kvantia and the three directions
How do Kvantia AI Office, Enia and Kvantia RIG-13 fit together?
They are three directions built by one company. Kvantia AI Office supervises an AI worker doing office work. Enia is a trainable transformer model, and Enia on device is its personal, local form. Kvantia RIG-13 is a local AI server for companies: the language model runs on your premises. It is available on request. The products stay distinct; the team, evaluation methods and deployment discipline are shared.
What is available today, and what is coming?
Kvantia AI Office is available for Windows and macOS, with a seven-day trial. Enia on device runs as a working prototype. Enia is in development. Kvantia RIG-13 is available on request. The roadmap shows the sequence, with one real status per item.
Who is behind Kvantia?
A small deep-tech team in Latvia. The legal entity is CODARS DESIGN, SIA. More on the about page.
Kvantia AI Office
What is Kvantia AI Office?
A supervision layer for an AI worker: a task inbox, routines, approval gates and a full audit trail. The AI only gets the capabilities a task needs and waits for your approval on risky actions. Explore AI Office.
How do I try it?
Choose the Solo or Office plan and start a seven-day trial on the AI Office pricing page. Detailed product questions are answered in the AI Office FAQ.
Enia and Enia on device
What does “trainable” actually mean for Enia?
Enia is a classic transformer model rebuilt from an open-source model base so it can keep being trained and specialised for a domain, instead of being a fixed snapshot you only prompt around. Before release we will name the base model and licence, state what we changed, and describe the method — continued training, fine-tuning or adapters — in plain terms. It is in development.
What is Enia on device?
A personal AI designed to run on a device you control, keep chosen context and learn only from what you teach it. The core runs today as a working prototype; persistent memory is in development. Explore Enia on device.
Does Enia on device send my conversations to Kvantia?
The intended architecture keeps conversation and personal-memory processing on the device. Account and licence data use the account service. See the data boundary for the details and their current status.
Does Enia on device work offline?
It is designed to. Conversation and personal memory are designed to run on the device, so the core interaction does not need a connection. Account and licence checks use the account service and need to connect from time to time; how often will be stated before early access opens.
Can I run Enia on device on more than one device?
Yes, that is the plan. Each personal AI lives on one device you choose, a desktop, a phone or a home server, with its own memory. Two devices mean two personal AIs. How additional devices are offered will be published before early access opens.
Can I move Enia on device to a new device?
Your account is not tied to the hardware, so it moves with you. Carrying a personal AI’s memory to the new device requires encrypted backup, which is on the roadmap.
What hardware will Enia on device need?
These are target requirements, not final ones: a desktop on macOS 13 or later, Windows 11 or Linux, with 8 GB RAM and about 4 GB of free disk; a phone on iOS 17 or later or Android 13 or later; or a home server running Docker with 8 GB RAM. The apps are in development, so the targets may change before early access.
Kvantia RIG-13
What is Kvantia RIG-13?
A local AI server for companies: the language model runs on your premises, and company data is not sent anywhere. It is available on request. Read about RIG-13 or contact us.
Early access and investing
What is the money for?
The next milestones: deeper product work on AI Office, model training and evaluation for Enia, early access for Enia on device, plus the team, licensed domain data and compute to do it. This is deep product and model work and carries real risk — the invest page sets out the evidence, milestones and risks.
How do I get early access or invest?
Join the early-access waitlist to be first into Enia and Enia on device as they open. Kvantia AI Office needs no waitlist, and RIG-13 is available on request through contact. To invest or pilot a model, use the invest page or contact us and we’ll follow up with the deck or an early-access plan.