The Atlant Platform

How does the Atlant Platform work?

Principle of Operation

Atlant is an enterprise-wide platform for knowledge acquisition and processing. To acquire knowledge, Atlant analyzes existing documents, extracts their content, and stores it in linked databases and knowledge graphs. For retrieval and further processing, Atlant provides users with chatbots, search forms, and other tools. Users can use these to ask questions in natural language to the Atlant platform, which answers them using language models (LLMs) and the previously created databases and knowledge graphs.

How does the Atlant platform work?

Structure & Procedure

In Source systems like SharePoint, Alfresco, or S3, your knowledge is stored in form of many, many documents.

With Workspaces you organize your knowledge by topic and assign access rights to users and groups.

With an automated Import process Atlant retrieves documents from the source systems and feed them into your workspace.

A configurable Indexer process in your workspace extracts your knowledge and stores it in specialized databases and knowledge graphs.

The retriever process in your workspace uses internal or external LLMs, in order to generate valid and verifiable answers from your queries.

To access the retriever, configure it for your workspace. Chatbots, Search interfaces or Agents.

What features does the Atlant platform offer?

Functions that we build for you

Documents & Formats

Work with the formats you are familiar with:

PDF, RTF, TXT
Word DOC, Excel, PowerPoint
Web: HTML, XML
PNG, JPEG, GIF, TIFF (with OCR)

Workspaces

Organize your knowledge in freely configurable workspaces. Each workspace can be configured with its own workflows for indexing and retrieval.
You control access using configurable user groups.

Text Reference

In the document view, you are shown precisely the text passages that formed the basis of the AI response.

CMS & DMS

Import your documents from leading Document Management Systems (DMS) or integrate directly with them:

Documentum OpenText
Alfresco
Nuxeo
CMIS Server
MS SharePoint

Solution Packages

Choose the solution package that best suits your application. Each solution package is tailored to your document type and requirements and is continuously updated to reflect the latest technological advancements.

File Systems

Import from all major file systems:

local file systems (Windows, Linux)
Amazon S3
Google Cloud, Google Drive
Azure BLOB
MS OneDrive

Private / Public LLMs

You can use internal (local, private) language models or public ones like OpenAI, Claude, Grok, Hugging Face, etc.
If your company or organization operates its own LLM program, it can usually be used in conjunction with Atlant. This way, your data remains within the company.

MCP Server

The Atlant platform supports the MCP (Model Context Protocol). It connects the Atlant platform to other AI services, external applications, and automated workflows.
(Magellan Edition)

Bots & Agents

On the Atlant platform, you can define your own bots and agents, separately for each workspace.

Automated Updating

When you add new documents to your document management system, or modify or delete existing documents, Atlant ensures that the changes are reflected in the answers to your questions within a short time.

Free LLMs

Within your workspaces, you can configure free LLMs, both public and private. Please note that with free public LLMs, your data is often used for further training.

Language Support

The Atlant platform supports all languages in which your documents are written and which the respective LLMs offer.
The user interface itself will be available in German and English. Additional languages can be configured upon request.

Take the Product Fit Test!

Learn more:

Frequently Asked Questions

Which Large Language Model (LLM) does Reelway use?

The Atlant platform is designed to work with virtually any commercially available LLM, including privately operated or company-owned LLMs.

What AI techniques does Reelway use?

Reelway uses LLMs as a basis but enhanced it with various well-known as well as modern techniques such as BM25 relevance ranking, hypothetical document embedding (HyDE) or model based evalution & optimization.

What is meant by “RAG”?

“RAG” is an abbreviation for “Retrieval Augmented Generation” and refers to a technique in which the query to a Learning Resource Manager (LLM) is enriched with text snippets from a document search. This ensures that the LLM has access to its local and current knowledge to answer the query.

Ask questions

Write to us if you would like to know more.

The form has been submitted successfully!

There has been some error while submitting the form. Please verify all form fields again.