AI services from Vienna, Austria
Enterprise AI where every answer shows its source
UNID AI is an AI services company based in Vienna, Austria. We make document archives and enterprise data usable by AI: every answer names its source, every number comes from your data, and if you want, everything runs on your own infrastructure.
- A source for every statement
- On-premise if required
- Based in Vienna
Animated sample view: an employee asks the AI assistant what the firm wrote about slope stabilisation for the Musterstadt project. Every statement in the answer cites file and page; one click opens the original page with the quote highlighted; then a Word draft is created in the company layout.
Products
Two products, one principle
Documents or tables: the AI works only with what is actually in your company, and shows where each answer comes from.
UNID AI Second Brain
The second brain for your company's knowledge
An AI knowledge assistant that makes your document archive searchable. Staff ask in the AI assistant they already use and get answers where every statement comes with its file and page.
Learn more →UNID AI Predict
Predictive and analytical AI on your enterprise data
An AI assistant for your company's tables: it answers questions with the query shown, checks data for anomalies and delivers predictions from a relational foundation model. On your own infrastructure if required.
Learn more →UNID AI Second Brain
Your archive, as a knowledge assistant.
Contracts, expert reports, quotes, minutes, scans too: loaded once, then asked in plain language, inside the AI assistant your team already uses.
- Every statement with file and page; one click opens the original
- Comparisons as tables, drafts as Word files in your own layout
- Login, permissions and a log of every access
UNID AI Predict
Predictions and analysis, straight from your data.
An AI assistant for your tables: it writes the query and shows it, checks for anomalies and delivers predictions from a relational foundation model.
- Numbers come from data and models, never from the language model
- You check predictions yourself in a backtest before you invest
- Fully on your infrastructure with a local model, if required
Animated sample view: a worklist shows the policies with the highest cancellation risk in the next 90 days, sorted by premium at risk. Clicking a policy shows the reasons from the data and the next best products. A backtest shows how the insurer checks the prediction themselves.
Animated sample view: a controller asks the AI assistant for the top five customers and the query is shown. The assistant then checks invoices for VAT errors and shows the ledger line behind each finding. Finally a prediction model lists customers likely to stop ordering, using a local language model.
Deployment
Where your data stays
You decide where the system runs. We state exactly which data goes where.
On your infrastructure
Everything runs in your data centre, with a local language model and no outside connection if you want.
On a controlled server
Archive, index and records sit on a dedicated server. Only the passages an answer needs go to the assistant.
With cloud models
If you want to use a cloud language model: with a processing agreement under Art. 28 GDPR and a documented data flow.
Approach
How we work
First call
We clarify goal, data and constraints. 30 minutes, no obligation.
Data check
We look at your documents or tables and tell you honestly what works and what does not.
Pilot with a criterion
A pilot with a measurable success criterion agreed up front.
Operation
Roll-out, operation and further development from one contracting party in Vienna.
Customers & partners
References only with permission.
We are currently running pilot projects with our first companies in Austria. We publish references only with our customers’ explicit approval. Until then, we name no names here.
Become a pilot partner →Insights
Enterprise AI, clearly explained.
- Use case
Churn early warning in insurance: data, method and verification
What data an insurance churn early warning needs, how the prediction is made and how an insurer checks for itself whether it beats the existing model.
- Comparison
Running AI in-house: on-premise, own server or cloud compared
On-premise, a dedicated server or the cloud: where enterprise AI can run and what each option means for data, model choice, cost and GDPR.
- Explained
Searching company knowledge with AI: what works and what doesn't
How companies make their document archive searchable with AI: which approaches are reliable, where the limits are and how answers are backed by sources.
FAQ
Frequently asked questions
What does UNID AI do?
Which companies is UNID AI for?
Does our data stay in the company?
Which AI assistants does it work with?
How does a pilot work?
How is UNID AI related to UNID Solutions?
See it working on your own material.
In a 30-minute call we show the products live and work out which use case makes sense for you first. No obligation, in German or English, in Vienna or online.
UNID AI · Leonard-Bernstein-Straße 10 · 1220 Vienna · +43 660 476 30 89