Financial Risk Profiler
Build risk profiler for your loan applicant.
Datanomers AI (NLP and ML) big data platform builds the use case: Financial Risk Profiler (FRP) to
- Provide leading-indicator credit risk information from the web, summarized in a dashboard, neatly categorized in multiple tabs
- Automate credit risk assessment for commercial underwriting
- Quicker, consistent and better loan outcomes
- Cloud-based solution with fee based on monthly transactions
Web data has latest info on credit, leading-indicator -> complements core financials.
How does FRP differentiate:
- Credit Bureaus web data delayed by manual screening, no longer fresh and latest
- Startups geared towards stock equities, hedge funds, asset managers -> provide financial information from limited number of curated web sites
- FRP provides financial information from a billion web sites (social media, news reports, govt registries, SME website, etc) -> very importantly, adds further value by extracting business intelligence from data to better predict loan outcomes -> intelligence made actionable
Vision for FRP: democratize access to capital, especially for SMEs -> complements the thin financials, with web data, to lend to more, without increasing risk.
Milestones for FRP: 5 banks (BMO, AmEx) and FinTechs (Fundation, Quarterspot) are customers.
Banks need FRP to lend to SMEs with confidence – loan origination and portfolio monitoring:
Underwriter Productivity: Get latest information from billions of web sites about credit risk, neatly categorized for ready consumption -> eliminates keyword-based search engines that hit you with spam -> 10-15% increase in underwriter productivity
Reduce Defaults: Extract predictive intelligence from textual data by combining it with financials -> customized ML model to reduce defaults 5-15%, lend more with same risk -> actionable intelligence
Very easy to trial, no bank IT effort -> transmit anonymized data, we return results
Easily deployable and scalable on the cloud (e.g. Azure VMs, Azure Load Balancer, etc.).