GenerAItor: Custom AI services without pro­gramming

Cost-effective, no-code gene­ra­tion of AI services opens up new fields of appli­ca­tion, eli­mi­nates typical problems asso­ciated with the avai­la­bi­lity of AI expertise and IT capa­ci­ties, and enables requi­re­ments to be imple­mented much more quickly.

Insiders presents GenerAItor, a new solution that allows the easy creation of cus­to­mized AI services for the clas­si­fi­ca­tion of documents and the automated extra­c­tion of data and facts from all types of documents – be they emails, chat messages, PDFs, scans, or other documents.

Business tran­sac­tions are often triggered by incoming documents, which pre­viously had to be processed manually, a laborious task. Automated pro­ces­sing using cost-effec­tively generated AI services therefore offers enormous potential for improved quality, speed, and cost savings. The intel­li­gent auto­ma­tion spe­cia­lists at Insiders have now developed a tool that allows such AI services to be created without pro­gramming knowledge or special AI skills. Since the large language models (LLM) used, among other things, do not require training, the newly generated services can be used imme­dia­tely. GenerAItor selects the best AI tech­no­lo­gies for each indi­vi­dual data pro­tec­tion requi­re­ment using a best-of-breed approach. This allows modern AI to be intro­duced into a wide range of business processes with low barriers to entry.

Data extra­c­tion defined with just a few clicks

The creation of an AI service always begins with the upload of an indi­vi­dual sample document. Regard­less of the structure, the AI already reco­gnizes poten­ti­ally inte­res­ting fields and suggests them for extra­c­tion. Users can easily add new fields with a click and com­mu­ni­cate their expec­ta­tions for the extra­c­tion result in natural language. In the insurance context, for example, they could say: “Assign the damage report to one of the areas of martens, stone chips, vandalism, or accidents.” Modern language models are now so advanced that they can even handle abs­trac­tions of facts. In the example, a damage event can be iden­ti­fied as rockfall, even if the specific terms “winds­hield” and “rock” do not appear in the descrip­tion. In addition, infor­ma­tion about the desired for­mat­ting of the results can be easily specified. Within minutes, a powerful service can be defined, tested, and deployed ready for pro­duc­tion. Ready-made AI services in the Insiders Mar­ket­place round off the range of services offered by Insiders Tech­no­lo­gies.

Product receives enthu­si­a­stic response

As part of its customer-centric inno­va­tion process, Insiders Tech­no­lo­gies has already gained extensive practical expe­ri­ence with the new GenerAItor in col­la­bo­ra­tion with selected customers and partners. Some partners are already using the solution pro­duc­tively to enhance their industry solutions with the capa­bi­li­ties of LLM.

Customers who are already using GenerAItor report a signi­fi­cant increase in the speed of imple­men­ting requi­re­ments since adopting the self-service model for gene­ra­ting AI services by the spe­cia­list depart­ments. „The potential of Insiders AI OvAItion, espe­ci­ally through the ability to integrate LLM, is huge. We currently have a to-do list of around 50 use cases that we want to automate and optimize with AI. The Insiders Cloud and the no-code GenerAItor make imple­men­ta­tion par­ti­cu­larly easy for us,“ explains Timo Kohler, Product Owner Input Manage­ment at the Baloise Insurance Group.