Knowledge base (RAG)
Make corporate knowledge accessible securely and easily
Benefits at a glance:
- Transparency: Corporate knowledge is accessible to the entire company and is quickly available.
- Efficiency: Quick retrieval of information improves work processes and increases employee productivity.
- Easy to use: By making it easier to obtain information, work processes are optimized and team productivity is promoted.
- Administrator control: Approval and administration of information remain centrally controllable.
- Data protection: GDPR-compliant solution with maximum protection, including business-critical data.
challenge
The knowledge available in companies is often hidden in complex file structures, extensive wikis or on intranet sites. To get the information you need, you usually have to have a certain idea of what you're looking for, how to start the search, and where the information you need is.
solution
nele.ai provides a solution that makes it easy to access corporate knowledge using knowledge databases and artificial intelligence (AI). The knowledge databases are the technical process “Retrieval Augmented Generation”/RAG. To use this option, you must first upload all relevant documents to the administration area of nele.ai by an administrator. Supported file types include EPUB, JSON, JSONL, PDF, XLSX, CSV, DOCX, and TXT. You can create various knowledge databases and assign them to individual users.
For example, questions such as “How do I apply for vacation?” , “What should be done if the delivery is damaged?” or “Who is responsible for conflicts within the team?” will be answered, provided that you have embedded the information beforehand.
The uploaded data is prepared in such a way that it can be combined with any AI model. They are embedded in secure servers in Germany. This process does not include training the AI, but the creation of so-called embeddings. When you select a knowledge base in nele.ai and type in a question or statement, your input is matched with the embedded information. The system then extracts the relevant section from the documents. The results are made available to the AI model along with your input or question, making the model better informed to answer your question.
Please note that when you select a knowledge base, only the information you have provided can be used. General information that is outside of the content you have provided cannot be answered.
You can access this knowledge directly via the nele.ai chat window. Select the appropriate knowledge base and ask your question. The system will then provide you with the appropriate answer. Internal data does not leave the safe area, as neither requests (prompts) nor results are used to train the selected AI provider.
GDPR-compliant document storage
Companies today must handle their data strictly in accordance with data protection regulations to ensure that sensitive information remains within the company's secure area.
The nele.ai platform securely stores all documents, prompts and results on servers in Germany. As a result, they are subject to the strict regulations of the General Data Protection Regulation (GDPR). In addition, there are contracts with all AI service providers that ensure that they do not use the data.
We also carry out regular data protection checks. There is, of course, an order processing contract (AVV) for all data processing by third parties. These measures guarantee the protection of personal and business-relevant information.
Using knowledge databases
In order to give employees access to the stored information, the “knowledge database” function must first be activated by the administrator.
Once approved, the corresponding knowledge databases can be created, documents uploaded and embedded. Employees can then access the information in these documents.
Employees enter a question, select the knowledge database and the AI only answers with the information contained within the documents. Data from the Internet is not used to answer.
Use case: Onboarding new employees
For example, you can improve your onboarding process with the help of AI. The onboarding process for new team members can thus be significantly simplified and accelerated. For this purpose, a separate knowledge database can be created, which contains all important familiarization information. The new team members get access to this database and can retrieve the information they need by asking a simple question. Questions such as “Are there opportunities for continuing education during my probationary period?” , “Under what circumstances can I take special leave?” , “What travel expenses are reimbursed?” can be easily clarified.
Another example of a service would be to help you choose a product by answering specific questions. For example, a question could be: “Which product is best suited to run at 1000 rpm in an environment with sub-zero temperatures?”
A similar example could be that a customer could need advice to select the right lubricant for a machine that works continuously at high temperatures. A corresponding request could be: 'Can you recommend a lubricant that will maintain its performance even when used continuously in an environment with temperatures above 200°C? '”
Employees no longer have to laboriously work their way through folders and text documents, but receive the information they are looking for directly.
synopsis
Secure and intelligent access to information is essential in corporate information management today. With its knowledge databases and advanced AI technologies, nele.ai offers the ideal conditions for this.
This mix creates great added value for companies. Organizations specifically rely on solutions such as the nele.ai knowledge database. As a result, they increase productivity and efficiency without neglecting security and data protection standards.
