Need / Problem:

Company managers wanted to see daily/weekly sales figures, analyse the performance of a particular store or product group. However, data analysts were generally needed for such reports. Since field teams or management did not know how to write SQL queries, there was slowness and dependency in accessing information.

Solution:

The AI-powered sales chatbot developed within BearCockpit recognised natural language questions from users and generated automatic SQL queries in the background.

For example:

‘What was the best-selling product in Istanbul branches last month?’, a graphical response was returned within seconds.

This chatbot worked as an “agent” within BearCockpit and connected directly to the sales database with the necessary API connections.

Benefit Provided:

1. Instant access to data without the need for technical knowledge.
2. Independent working of management and field teams.
3. Time saving in decision making processes.

Need / Problem:

Aircraft maintenance teams were wasting time searching through technical documents (e.g. maintenance manuals, EASA/FAA guidelines). Moreover, these documents were in different languages (English, Turkish, French, etc.) and it was difficult for the teams to understand each document. It was difficult to get the right information at critical moments.

Solution:

All technical documents were processed with AutoRAG and entered into a vector database. Users were able to make queries in natural language and in different languages. In addition, technical constraints were dynamically extracted from a separate database to which the system is connected.

Benefit Provided:

1. Technical staff directly accessed the information without searching through the documents.
2. Language barrier was eliminated with multilingual support.
3. Operational errors were minimised and maintenance times were shortened.

Need / Problem:

There were hundreds of pages of legislation, manuals and internal procedure documents within the company. Employees were spending long periods of time to find the right information from these documents. In addition, the reliability of the answers received from the document was being questioned.

Solution:

With AutoRAG, all documents were divided into parts and indexed according to semantic meaning. When the user asked a question, the system generated answers by referencing only the correct parts. Moreover, these answers were shown with the references in the relevant document (verifiable answers).

Benefit Provided:

1. Information retrieval time reduced from minutes to seconds.
2. User confidence increased because the documents on which the answers are based are clearly shown.
3. Critical productivity gains for internal audit and regulatory teams.

Need / Problem:

The control of the serial numbers laser printed on the products coming out of the production line was done manually. This took time and created a process that was open to human error. Quality control processes were progressing slowly.

Solution:

The image processing module integrated into BearCockpit analysed images from cameras on the production line. The serial number on the product was read with OCR (optical character recognition) and matched with the correct reference number in the system. Defective prints were detected immediately.

Benefit Achieved:

1. The quality control process was fully automated.
2. The risk of faulty products reaching the customer has been eliminated.
3. The process became 24/7 traceable and reportable.

Need / Problem:

Although patients or healthcare professionals could access the package leaflet by scanning QR codes on medicine boxes, it was difficult to find the information they were looking for among these documents.

For example:

‘Can a breastfeeding person use this medicine?’, ‘What are the side effects?’
could not be answered directly.

Solution:

Prospectus documents were processed with AutoRAG. Users scanned the QR code and started asking questions in natural language on the chat screen. The system extracted the relevant information from the correct paragraph and presented it in an understandable way.

Benefit Provided:

1. Users were able to get direct answers to their questions without reading the document.
2. Access to information became easier for both healthcare professionals and patients.
3. The number of users accessing the right information increased.