- Home
- On-premises AI
On-premises AI: your enterprise AI in your own network
With CGS Assist, you bring your employees’ use of AI out of private accounts and third-party clouds into your own infrastructure. CGS deliberately offers no SaaS and no hosting: you run the platform yourself.
Runs in your environment, one Linux server is enough
CGS Assist is set up on a Linux server in your data centre or in your own cloud. The platform, documents and logs stay in your network. With a locally operated model, this also applies to all queries.
Your dedicated contact at CGS handles the setup: connecting the model, linking data sources, creating roles.
Bring your own LLM
CGS Assist is vendor-independent. You use local open-source models such as Llama or connected services such as OpenAI and can switch at any time without anything changing for your teams.
You choose the language model in the chat, per use case and per agent. Sensitive data stays entirely with you when you use a locally operated model. If other tasks use a cloud service, their queries are sent to that service and are subject to its terms.
Built for production
An AI pilot is quickly demonstrated. What matters is what comes next:
- Roles and permissions for use cases, documents, models and agents
- Audit logs of changes, administration, use cases and agents
- Central administration of users, token limits, use cases and licences
- Approvals, limits and policies that stand up to scrutiny by auditors and regulators
The alternative to the big SaaS platforms
Frequently asked questions
Where is our data stored?
Does data leave our company when we use a language model?
Book your demo directly in the calendar
Pick a free slot. When booking, tell us about a process from your daily work, and we will show you how CGS Assist handles it in your infrastructure.
- Duration45 minutes
- LocationOnline via Microsoft Teams, you receive the link with the confirmation
- WhoYour future contact at CGS, not a call centre