The AI layer as a service
Your data, our hardware, your rules.
We build and run for you what we work on ourselves: local models, one gateway with rules, answers over your knowledge, and agents that reach into your systems under control. Not a connector to the cloud, but a layer next to your data.
What the service includes
Six parts that hold together
Running local models
Open-source models on GPUs in our data centres in the Czech Republic. Model choice per task, load and capacity monitoring.
A gateway with rules
One OpenAI-compatible endpoint for all your tools. Rules by data class, limits and an audit log.
Answers over your knowledge
Open WebUI with RAG over your documentation, wiki and projects. Role-based access.
Connectors to your systems
MCP servers for GitLab, documentation and internal APIs through service accounts with a defined scope.
Custom agents
Agents over your mail, tasks or tickets on the same principles as our case studies: the model proposes, code decides, a person approves.
Governance in writing
Data classification, cloud rules and responsibilities written down at the start. Documentation for supplier assessment.
How to start
From a conversation to the first agent in production
- 01
Mapping
Where your data lives, which processes slow you down and what of it may leave your network.
- 02
Classification and rules
We write down data classes and model rules. That is the foundation, not a formality.
- 03
Pilot of one agent
One task, a measurable result, a human in the loop. Typically mail or ticket triage, or answers over documentation.
- 04
Operation and growth
The agent in production, quality measurement, and further tasks where the benefit is greatest.
Inspiration
What already runs at Nux
Each case study is a pattern that can be carried over to your systems.
IMAP triage
Sorting company mail and drafting replies
In the morning it is clear what is on fire. The agent sorts the NOC, CSIRT, billing and sales mailboxes and drafts replies to selected messages. It cannot send them.
GitLab triage
A daily issue digest for developers and teams
Every morning, what changed in GitLab and what is expected of whom. Code builds the sections; a local model only adds summaries. Client data never sees the cloud, and code enforces it.
Day digest
A morning overview from company systems
One e-mail before the day starts: calendar, tasks, mail, GitLab and Mattermost. The agent calls no model itself; it assembles the output of the other agents.
Delegator
Assigning tasks in plain text in chat
You tell a Mattermost bot what needs doing and who should do it. The model proposes a task, code checks the assignee and deadline, you confirm with a thumbs-up. Then the bot watches how the task goes.
Expense hub
Re-billing company expenses
Statements for fuel, phones, insurance and internet arrive by e-mail and end up as an overview for each holder. No model: the parsers are deterministic and the totals match to the cent.
FAQ
Before you get in touch
How much does it cost?
We set the price after an initial assessment, based on scope: how many tasks, how large the models and how much GPU capacity it needs. Routine tasks on local models are not paid per token.
Can we use cloud models too?
Yes, if your data rules allow it. The gateway lets them through only for the data classes you have allowed, and records every call.
What if a local model is not enough?
Then we say so and measure it. Either we break the task into smaller ones, or we choose a larger model for it, local or cloud if the data allows.
No-obligation consultation
Let’s find the first job for AI in your company.
An hour’s conversation about where your data lives, which processes slow you down and what of it may leave your network. You leave with a proposal for a first pilot, not a tool demo.
Built and is accountable for the AI layer at Nux
Radek Růžička
Managing Director and CTO, Nux s.r.o.