AI in Microsoft Dynamics 365

To assess damage cases, our client relies on a PowerApp for mobile image capture. From these images, AI generates an initial damage report. This automatically generated damage report can then be reviewed, refined, and corrected if necessary by the responsible employee. A further AI-based analysis follows in a subsequent step. Using text classification, the cause of the damage and the necessary work to be performed are identified from the finalized report.

• Simplified work on the construction site – no manual text entry
• Mobile use – user-friendly with PowerApp on tablet and smartphone
• Quick help – automatic identification of the next tasks
• Better data quality – structured data and consistent wording

Time-consuming, manual research is still routine in many purchasing processes. With just one click on the item card in the ERP, Copilot launches an automatic web search based on the respective item keywords. Relevant online shops are identified and can then be selected specifically. In the next step, Copilot searches these shops for suitable products and prices and presents the results in a clear, structured list.

• Real time savings – automated research instead of manual searches
• Easier decision-making – clear and comparable presentation of products and prices
• Current data through online search – ideal in fast-moving markets.

Screwdriver or bolt cutter? Stainless steel or V2A or simply rustproof? Damp room or IP44? Bucket or tub? People often mean the same thing but use different terms. Missing search results often lead to abandoned purchases and thus lost revenue. Employees' stress levels rise because searches have to be repeated with different terms before an item is actually found.

KUMAVISION's AI-supported semantic search finds similarities in terms of quality, purpose, meaning, etc., like an experienced salesperson who thinks ahead.

• Versatile use in price lists, product databases and online shops
• Higher conversion rates and revenues through better hits
• Less stress or frustration for employees and customers

A customer with many suppliers receives a large number of order confirmations daily in a wide variety of formats: PDFs, emails, and sometimes scans. Employees had to manually search for the relevant data. The KUMAVISION AI solution extracts the information and transfers it directly to the ERP system, where the next steps, such as matching it with the order, begin immediately.

• Measurable improvement in data quality
• Significant time savings and relief in day-to-day operations
• Easy scaling of the business during peak order periods, etc.

A KUMAVISION customer manages approximately 30.000 items and regularly receives reseller inquiries with varying descriptions. Previously, the process of matching these inquiries to the relevant product range was slow and prone to errors.

The new AI prototype uses a vector database for semantic search, recognizes similarities in meaning and automatically suggests suitable articles - even with highly varying descriptions.

• A 96% success rate (Top 10) proves the reliability of the AI ​​solution.
• The AI-generated results were successfully checked against a reference list.
• Shorter processing times
• Fewer errors and improved service quality

A medical technology company used KUMAVISION medtec to validate its ERP processes.365-Package including complete process list and GAMP5 risk analysis guide.

The team used AI-based methods to assess risks automatically and accurately. Existing data was analyzed using targeted prompts. The risk analysis is now integrated into the validation package and ready for immediate use by customers.

• Reliable process identification
• Structured risk allocation
• Real time savings and better data during validation