Work · Scienaptic AI

Land and expand: taking a decisioning platform from $2M to $10M

Platform revenue grew fivefold while revenue per client rose 50% and client retention rose 35%. First-time client go-live fell from 150 days to 30.

Context

Scienaptic sold AI credit decisioning to more than 100 lenders: mostly US credit unions, plus auto and specialised lenders, fintechs, and banks and NBFCs in India. It ran as multi-tenant SaaS in two environments, US-hosted and India-hosted, for data residency. Most revenue came from core decisioning, and bringing a new client live took about 150 days of bespoke implementation.

The hard call

Expand before chasing new logos. With a fixed team, every roadmap slot spent on the existing base was one not spent on winning new clients. I put the weight on products that 100+ lenders already trusting the platform could add. New logos still came, partly through those same products.

What I did

• Built iCUE on top of decisioning and cross-sold it into the base.
• Opened MSME as a new asset class.
• Replaced bespoke implementation with a self-serve decisioning framework.
• Added backtesting, audit traceability and champion-challenger routing to the business rules engine, so a credit strategy change for a live client went from 90 days to 1.
• Launched LLM features early as a paid differentiator.

My role

I owned the product roadmap and led the suite expansion and cross-sell, reporting to the COO and working closely with the founder and CEO, alongside engineering, data science, implementation and sales.

Results

What I take from it

Your cheapest growth is the client who already trusts you. Give them the next product they need and make it easy to switch on.

Published 19 Oct 2026

Next case: Making smallholder loans investable: Kenya's first private-sector agricultural securitisation