Work

Stop custom builds: said no to per-client models despite near-term revenue risk, in favour of configurable products.

Expand before chasing new logos: weight the roadmap toward products the existing 100+ lenders could add.

Coach first, lend second: treat a decline as the start of a plan, not the end of the funnel.

Start LLMs at monitoring booked loans, not at application decisions, and bundle three tools into one module.

Built ki catalyst as a reusable product before the first deal, not one-off analytics for one originator.

One platform, not a second product: MSME as an asset class on the shared consumer platform.