About

Karthik Velu is a product leader in lending, credit and AI. Over 14 years he has built the infrastructure lenders run on (scoring, origination, monitoring and capital markets) for banks, credit unions and NBFCs in the US, India and East Africa. He is Senior Director of Product Management at Kaleidofin. Before that, as Director of Product at Scienaptic AI, he took a credit decisioning platform from $2M to $10M in revenue. He holds an MBA from IIM Bangalore and has cleared CFA Level 2.

The longer version

I've spent fourteen years around credit: how it gets underwritten, priced, originated, monitored and funded. Most of that time has gone into the infrastructure lenders run on, for borrowers most credit systems can't see.

Today I lead product, design and marketing at Kaleidofin, across KPL, which builds risk technology for lenders, and KCPL, the group's NBFC. When I joined in May 2025 the company built scoring models client by client. With a team of nine we turned that into four platform products (ki score, ki monitor, ki catalyst and Ki Originate) and built Dada Biashara, a GenAI financial coach on WhatsApp for women micro-entrepreneurs in Kenya. The same infrastructure ran Kenya's first private-sector agricultural securitisation.

Before that I was Director of Product at Scienaptic AI, owning the roadmap for a multi-tenant credit decisioning platform used by more than 100 lenders, mostly US credit unions. Platform revenue went from $2M to $10M, and we put LLMs into production early, starting where they were safe. At Yubi I automated securitisation end to end on India's largest debt marketplace. Earlier I worked on Wio's payments with Publicis Sapient, real estate lending at Barclays and supply chain finance at Bank of Baroda, founded a small home-automation startup, and started out as a programmer at Cognizant.

The part of the job I find most interesting is deciding what the platform absorbs, so that everything built on top of it gets cheaper.

What I believe about product

• Custom work wins the first deal and caps the next hundred.
• Put a new technology where its mistakes are cheap and its value is obvious, then move closer to the decision.
• A decline is a product moment, not the end of the funnel.
• Your cheapest growth is the client who already trusts you.
• Where capital is scarce, the product is trust.

Timeline

The full outcomes-led version →