Major project Banking & Fintech Case Studio

Smart payment router with live learning

Inspired by publicly reported work at Razorpay

Send each payment to the gateway most likely to succeed, learn from results, and fail over when a gateway degrades.

The brief

Razorpay has described Optimizer, which routes each payment to the gateway most likely to succeed using models trained on past transactions, and falls back automatically when a gateway's success rate drops. Build a smart router of your own: a simulator of gateways whose success rates change, a model that predicts success per payment, a bandit that learns live, and automatic failover. Good looks like: success-rate uplift over fixed and round-robin routing in the simulator, quick reaction to an outage, cost-aware routing, and a dashboard.

Read the case first

What Razorpay has said in public about this kind of work, on its own pages:

Make it your own

Route something else between providers that fail: SMS gateways for OTPs, cloud regions for an app, or courier partners for parcels. Keep the same learning and failover, and change what success and cost mean.

AITeky has no link with Razorpay, which has not reviewed or endorsed this brief. The brief only draws on what the company has published, uses open data rather than its data, and is a learning exercise: don't present your project as Razorpay's work or system.

What you will build

Build a gateway simulator with changing success rates, outages and fees
Train a model that predicts success per payment and gateway
Add a bandit that learns from live results
Fail over automatically when a gateway degrades
Compare uplift and cost against fixed and round-robin routing
Build a routing dashboard

Free, open data to build it with

It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.