Customer
The Wirtschafts- und Infrastrukturbank Hessen (WIBank), the development bank of the State of Hesse and part of Landesbank Hessen Thüringen plays a key role in infrastructure, economic, social, and housing development initiatives. It supports a wide range of customers in Hesse with financing, advice and extensive exchange and networking opportunities.
Initial situation
Finding appropriate funding can be challenging because of the diverse range of funding programs available and the corresponding funding requirements. WIBank offers an extensive array of consulting services aimed at uncovering the untapped funding opportunities within your project. Up to this point, the responsibility for this task has been held by the business development consultancy, which operates through regional offices and a telephone team based in Wiesbaden. This comprehensive service is currently being enhanced by a smart, digital voice assistant known as the WIBot.
The goal of WIBot is to create a connection between individuals in need of guidance and personalized advice from a real person. This allows interested parties to gather preliminary information and subsequently delve into the specifics of their project and potential funding during a face-to-face meeting. The bot provides an interface to „Wirtschaftsförderberatung Hessen“ located at WIBank, offering a comprehensive overview of all available funding opportunities. „However, technology is meant to serve as a tool to assist people, not to take their place. The WIBot therefore checks the facts, while the consulting team can creatively assess the individuality of a project. The only path to achieve genuine multiple benefits for individuals seeking guidance is through this clever blending of both strengths,“ summarizes Dr. Johannes Loheide, WIBank, Head of Department Fundamental Questions.
Challenge
The aim was to intelligently link VÖB-Service‘s vast funding database with the needs of the Hessian Funding Finder while ensuring practicality and structure were not sacrificed. Additionally, it was essential to uphold access to all funding programs to create a robust query framework. However, WIBank‘s API used a proprietary schema, requiring msg to initially transform the data into a format that could be processed. After overcoming this challenge, attention shifted to the particular needs of the Funding Finder integrated into the WIBank homepage. Filters had to limit the search and be customizable as required. It was essential to restrict the search results to properly prepare the Funding Finder and uphold its role as a guide in the funding jungle. In addition, querying the funding database should ensure that federal and EU funding is taken into account alongside the programs of the state of Hesse. This played a pivotal role in selecting a funding database provider.
Another primary concern for all stakeholders - msg, WIBank and VÖB-Service - was the content‘s quality and technical implementation. At the same time, gathering user feedback and statistics was crucial for continuous improvement.
Solution
msg‘s powerful Neo Enterprise Assistant Platform (NEAP) was utilized in tight collaboration with VÖB-Service to address the challenges faced by WIBank in creating the Funding Finder bot. Each partner brought their own expertise to the table: msg provided the bot technology, VÖB-Service provided the database and WIBank contributed their consulting experience. The bot provides a comprehensive overview of the various funding options available to businesses in Hesse.
First, the API of the VÖB-Service was connected to the Neo Enterprise Assistant Platform. The realization and implementation of NEAP in WIBank‘s web interface came next. As part of a comprehensive test, specific filters were defined for the type of funding, such as the reason for funding, number of employees, etc., and the regional search. Filters are converted into internal codes and IDs to ensure accurate search results, which are then passed on to the business logic for processing funding queries. Additional filters restrict the number of results. If further search results are relevant, these are collected in a PDF document and made available for export. If further advice is required after using the bot, the results of the bot search are sent as a copy to Wirtschaftsförderberatung Hessen. WIBank was particularly concerned with adhering to strict data protection standards, which is guaranteed through the anonymization of data. During the personal consultation, the advisor only accesses the protected data via a reference number. At the end of the inquiry process, those seeking advice can use a feedback feature to rate and comment on the process. This function is important for continuous optimization of the pilot project.
Outlook
The Funding Finder is presently undergoing a pilot phase to assess its acceptance. However, in view of the numerous advantages that the Neo-Bot offers both banks and funding applicants, a medium-term upgrade to a large language model is already planned. This aims to improve digital advice and further reduce the workload on staff.
msg and WIBank continuously monitor the number of inquiries and success rates in order to continually improve the bot. Marcus Kindermann and Claudia Ungeheuer stress the importance of viewing the Funding Finder as a multiplier for other funding institutions, which can result in wider application and greater benefits for the entire industry. „An intelligent (voice) assistant like this can support many other companies in Hesse,“ says Claudia Ungeheuer, responsible for public relations and consulting at WIBank