OptoPred (Utilization optimization for SMEs and self-employed through predictive matching) is a joint research project with Kassel University funded by the land of Hessen as part of the program Distr@l.
The volatility in the project market triggered by the corona virus is a major challenge for medium-sized companies: Many companies have to find new projects for numerous employees at the same time. The long duration of the review, application and selection processes leads to liquidity bottlenecks threatening the existence of the company, although numerous projects are still advertised. Economic potential remains untapped. It is therefore particularly important right now to seamlessly bring the supply and demand sides together through efficient and transparent matchmaking. Previous solutions on the market cannot do this for three reasons:
1. Clients are often uncertain about the credibility of the stated expertise of a contractor and are reluctant to award contracts.
2. You only consider a few factors that are necessary for a suitable and relevant matching. So it is mostly a matter of pure keyword matching, in which qualitative factors (e.g. the duration of experiences) and soft data are not taken into account.
3. Existing platforms are closed systems that only offer their own projects and function as classic marketplaces, while Lyncronize as an agent constantly analyzes all of these platforms for suitable projects, thus making registration on dozen different platforms superfluous.
The OptoPred project aims to fill these projects more quickly and to help small businesses and the self-employed to keep their workload high. For this purpose, the applicants develop an AI-based prediction model, which specifically predicts the success probabilities of project staffing. The model mainly takes the following dimensions into account:
• Quality indicators of profiles, esp.
• credible evidence of skills and competencies
• Qualitative information in the tenders
Self-employed and SMEs receive support from this model in order to control their profile design and competence development in an informed manner and to improve their prospects of success in acquisition. By integrating the prediction model into the Lyncronize matching algorithm, it is expected to achieve a significantly higher matching quality and enable projects to be placed more quickly. The applicants are not aware of any approach in science or practice that could provide transparent qualitative matching to the extent described.
The partner in the research project is the business informatics department of University of Kassel from Prof. Dr. Jan Marco Leimeister . The department researches design-oriented and with a high practical focus on the topics of innovation management, crowdsourcing, hybrid intelligence and collaboration engineering. In the ETH Zurich research ranking, Prof. Leimeister is in fourth place among the most research-oriented business administration chairs in German-speaking countries.