As asset-based finance expands beyond traditional lending into sectors such as data centers, buy-now-pay-later receivables, cell towers, and royalty-backed assets, investors are demanding more sophisticated tools to assess risk, model cash flows, and meet evolving regulatory requirements. In this interview with Fintech Business Asia, Marco Masotto, Head of Product at Cardo AI, explains how the company’s new cash flow modeling engine is addressing the shortcomings of legacy platforms, integrating real-time market data, and leveraging artificial intelligence to enhance decision-making across structured finance and private credit markets.

Q: Cardo AI's new cash flow modeling engine has been positioned as a modern alternative to legacy structured-finance platforms. What were the biggest limitations you observed in traditional tools, and why did you believe the asset-based finance market needed a fundamentally new approach?

A: Legacy platforms were built for a narrow set of standardised asset classes and haven't meaningfully evolved since. The core failures were inflexible waterfall logic that couldn't handle bespoke structures without expensive custom development, focus on static portfolio that made most simulation stale in case of dynamic portfolios, no live market data integration, and heavy reliance on offline spreadsheets for scenario work. Analysts were spending most of their time on data plumbing rather than credit analysis. We built from first principles because patching legacy architecture only goes so far. The market needed a platform where cash flow logic, triggers, and waterfall mechanics can be configured dynamically, and where the entire workflow from data ingestion to scenario output lives in one auditable environment. That’s what we built at Cardo AI.

Q: Asset-based finance is rapidly expanding into non-traditional collateral classes such as data center financing, buy-now-pay-later receivables, cell tower assets, and royalty-backed structures. How is this evolution reshaping investor requirements, and what challenges does it create for risk modeling and portfolio management? 

A: Non-traditional collateral (data centers, BNPL receivables, cell towers, royalties…) demands a multidimensional risk lens that generic structured finance templates simply weren't designed to handle. A data center deal, for example, requires modeling power infrastructure constraints and hyperscalers concentration risk, not just borrower financials. BNPL vintages behave very differently from traditional consumer credit and require granular cohort analysis. The deeper challenge is portfolio-level consistency. Investors increasingly need deal-level flexibility alongside standardised portfolio outputs.

Q: A key differentiator of the new platform is the integration of live Bloomberg rate curves for benchmarks such as SOFR, SONIA, and EURIBOR. How does access to real-time market data improve decision-making compared with traditional static cash-flow models and spreadsheet-based analysis? 

A: In a spreadsheet-driven workflow, rate assumptions are updated manually and infrequently, meaning models almost always run on stale data. In recent years, the gap between last week's curve and today's market can produce materially different pricing and stress outcomes. Live Bloomberg integration eliminates that latency entirely. Every scenario reflects the actual forward curve at the moment of analysis. It also enforces consistency across teams; everyone works from the same live inputs rather than individually maintained spreadsheets with different vintage assumptions.

Q: Insurance companies have become major investors in private credit and structured finance. How does Cardo AI's new modeling capability help insurers meet increasingly complex regulatory, accounting, and stress-testing requirements while improving transparency across investment portfolios?

A: Insurers face the most demanding analytical requirements in this market. Under Solvency II and equivalent frameworks, capital treatment depends on the quality of underlying cash flow models, and IFRS 9 requires dynamic, forward-looking expected credit loss calculations rather than historical averages. Our platform allows the same modeling engine that drives investment analysis to generate regulatory and accounting outputs directly, removing the manual reconciliation between analytical and compliance workflows. Equally important, insurers can maintain an independently modelled view of underlying collateral performance rather than relying solely on servicer reporting.

Q: Artificial intelligence is increasingly being embedded across financial analytics and risk-management platforms. How do you see AI enhancing cash-flow forecasting, scenario analysis, and asset-based finance operations over the next few years, and what role will automation play in supporting investment professionals rather than replacing them? 

A: The most immediate impact of AI is simply getting analysts to answer faster. Today, a lot of time is lost to repetitive work, cleaning data, extracting information from documents, preparing models. AI will handle most of that automatically, freeing teams to focus on what matters. On the forecasting side, machine learning can surface patterns in loan performance data that traditional models tend to miss, especially for newer asset classes where there isn't much historical precedent to draw from. That said, AI won't replace investment professionals. The judgment calls (how to underwrite a credit, how to negotiate a structure, how much risk to take on) require experience and context that no model can replicate. What AI does is take the grunt work off the desk so that professionals can spend more time on those decisions. At Cardo AI, we think of AI as something that makes good analysts more effective, not something that substitutes for them.