Chat AI is a privacy-oriented AI chatbot interface for large language models (LLMs). It supports retrieval-augmented generation (RAG), custom system prompts, and reasoning models. Chat histories are only stored on the client side.
Description
Chat AI (an external service provided by KISSKI) is an AI chatbot interface that does not store your chat history on the server side. This service makes it possible to use various large language models (LLMs), including their own system prompts, free of charge in a user-friendly chat interface.
It also offers additional services:
- Hosting your own language models
- Fine-tuning LLMs on your data
- Retrieval-Augmented Generation (RAG), an effective and cost-efficient technique to give an LLM constant access to the knowledge in your documents
Applications
- Technical support for employees: Fast answers to questions about machines, production processes, or software
- Productivity tool: Drafting and summarizing emails or documentation.
- Production planning: Assistance with material requirements planning and scheduling
- Customer service & supplier support: Answering product questions, technical specifications, or delivery dates based on RAG information
- Multilingual communication: Immediate translation and response to inquiries from international customers
How to get access
To use Chat AI, you must register via AcademicID. Registration is possible free of charge.
There is a basic limit for using the chat service:
- 10 API accesses per minute
- 200 API accesses per hour
- 400 API accesses per day
- 3000 API accesses per month
This limit can be increased upon request.
To use this service, go to:
Service provided by
Contact
Jonathan Decker
GWDG
Also:
Ali Doost Hosseini
GWDG
Jakob Hördt
GWDG
