Dubai-based Commercial Bank International (CBI) has announced its investment in zypl.ai, a fintech startup pioneering the application of generative AI in lending. Since launching sales, zypl.ai has crossed $1 million in annual recurring revenue at a more than 20% compounded monthly growth rate, and it is already operationally breaking even.
In addition to global expansion with major financial institutions across continents, the fintech is eyeing microfinance in Southeast Asia and credit unions in North America in particular, along with a foray into underwriting in the insurance industry.
zypl.ai’s flagship software – zypl.score – enables financial institutions to optimise credit decision models across retail and SME portfolios by generating synthetic data. Over 35 banks across 12 markets in Eurasia have deployed zypl.ai’s proprietary AI-as-a-service software, underwriting >$100 million of credit portfolio at minimal default rates. Zypl.ai has built proprietary generative adversarial networks (GAN) on a cumulative dataset of over 54 million traditional and alternative data points to power its credit decisions. This unique approach provides financial institutions with macro-resilient decision-making on dynamically trained AI models that factor in changing macroeconomic cycles.
When deploying zypl.ai, enterprise clients can fully customize zypl.score to their risk appetite with the goal of reflecting competitive market advantages. Through on-premises deployment, zypl.ai ensures full data privacy and confidentiality. Thanks to its interpretability features with embedded ‘weight- of-evidence’ for each credit decision, zypl.score is fully compliant with emerging regulatory conditions that require explainability of ‘black-box’ credit models.
CBI and zypl.ai initially launched a proof of concept in the summer of 2023, conducting a backtest of zypl’s synthetic AI algorithms on CBI’s historical retail dataset. The results were convincing: with minimal application data lacking credit history information, zypl.ai could reduce default rates by two-thirds whilst retaining 90% of non-default customers. Following the positive outcome, CBI entered into a commercial pilot of zypl.score in the fall of last year, with an expanded scope of retail decision model that included risk rating and Large Language Module (LLM) – driven decision interpretation capabilities. Currently, the Bank has proceeded to fully deploy zypl.score in its retail lending through a software licensing agreement with zypl.ai.
Commenting on zypl.ai, Ali Sultan Rakkad Al Amri, CEO of CBI, said “Our partnership with zypl.ai, an innovative AI platform, signifies a strategic entry into using AI implementation for advanced credit scoring. By harnessing zypl.ai’s synthetic data AI powered algorithms, we’re confident we can further enhance our operational efficiencies to ensure a swift and seamless lending experience. Our work and investment with zypl,ai aligns well with our commitment to supporting innovation in the UAE and with our strategic ambition to partner with only the most innovative partners in the financial technology space. We look forward to supporting zypl.ai’s growth both regionally and internationally.”
