While the core of Subba provides a robust, fault‑tolerant recurring billing engine on top of Nomba’s rails, our vision extends far beyond basic subscription management. We’re building the first billing infrastructure localized for the realities of the African economy, augmented by Artificial Intelligence.

Here is our roadmap for the features we’re building next.


1) AI‑Powered “Smart Dunning” & Predictive Retries

Traditional billing engines retry failed payments on a rigid, hardcoded schedule (e.g., 1 day, 3 days, 7 days). This leads to high failure rates in emerging markets because it ignores the customer’s actual liquidity patterns.

The Subba approach

We are developing a Machine Learning pipeline that analyzes a customer’s transaction history to determine their Peak Liquidity Window (e.g., detecting that a user usually receives salary deposits on the 26th around 8:00 AM).

Instead of blindly retrying a failed charge, our AI schedules the retry exactly when the transaction is statistically most likely to succeed. Furthermore, we map Nomba’s specific decline error codes to our logic:


2) AI Churn Prediction

SaaS developers are usually blind to churn until the user actually cancels.

The Subba approach

By analyzing payment patterns—such as frequent card declines, delayed virtual account funding, or downgrading plans—Subba will automatically flag users in the merchant dashboard with a “High Churn Risk” badge. This enables merchants to trigger preemptive retention offers or personalized outreach before the customer officially cancels.


3) Natural Language Revenue Analytics (LLM)

Dashboards traditionally restrict merchants to pre‑built, rigid charts.

The Subba approach

We are introducing an AI‑driven search bar in the Tenant Dashboard. A developer can type: