The subject of the tender is a contract for the following service: AI-based data extraction from credit documents for the digital WRF credit process. As the development bank for North Rhine-Westphalia, NRW.BANK supports the region in its structural and economic policy tasks. It acts under a public mandate, in a competition-neutral way, and employs the full range of banking development products—from traditional loans to tailored advice. The three support areas of NRW.BANK are "Economy", "Housing", and "Infrastructure/Communities". Sustainability is a central guiding principle. NRW.BANK promotes the necessary transformation of North Rhine-Westphalia towards a low-emission and climate-resilient economy and living environment.
The "Public Housing Support" sector is one of the key support segments in the "Housing" field. The target group comprises owner-occupiers or tenants within defined income limits. For owner-occupiers, the support is provided directly within the framework of property ownership support, while tenants are indirectly supported through loans to investors in public rental housing support.
In the context of rental housing support, NRW.BANK acts as a lender. This includes the entire chain from credit analysis to credit processing, property valuation, and problem loan processing. In new business, the credit process is preceded by the examination of eligibility and demand criteria by municipal approval authorities. NRW.BANK makes the necessary credit decision in the 1-vote procedure following this third-party initiation and acts in the housing support sector according to MaRisk as "market follow-up". This is followed in existing business by credit management and processing, including annual credit monitoring throughout the duration of the business relationship. NRW.BANK's portfolio in rental housing support amounts to approximately €15 billion and includes about 8,000 rental investors in a heterogeneous structure (especially private rental investors, traditional housing industry with municipal or cooperative background or capital market reference, investors with charitable background). Each year, NRW.BANK prepares around 2,500 credit templates (= aggregation to risk units) in new business and as part of the annual credit monitoring.
In property ownership support, new business is completely handled by municipal approval authorities without a credit assessment by NRW.BANK. The bank handles credit management and processing in new and existing business up to problem loan processing. The existing portfolio comprises around €2 billion with over 40,000 customers.
LOT-0001
NRW.BANK - AI-Based Data Extraction from Credit Documents for the Digital WRF Credit Process.
The operational processes of public housing support are a central success factor in both new and existing business. Currently, the process is very labor-intensive: especially reviewing customer documents and manually transferring data to the rating and evaluation system bind significant capacities. Given the continuously high demand for subsidies and increasing case numbers, process optimization is required. To make the analysis process future-proof and achieve significant efficiency gains, a technical solution should therefore be implemented. Both on-premises and cloud-based deployment models are permissible, provided the respective requirements for data protection, IT security, and integration into the existing system landscape are met.
The contract has a basic term that ends 4 years after Go.Live. The contract can be extended by NRW.BANK three times, each for 1 year.