Bank routing and AI

Every card,
to the right bank.

In-store POS routing is fully separate from virtual POS: its own rules, its own cost table, its own bank health and its own AI model.

Strategy

BALANCEDCOST_FIRSTAUTONOMOUS_AIADAPTIVE_AI
  • Bank A · TID 1chosen
  • Bank A · TID 2limit reached
  • Bank B · TID 1briefly closed

The rule picks the bank; if it has several TIDs, cost, limit and health decide.

What's included

From rules to AI.

  • Rules

    Amount, currency, instalments, store, register, device, day and hour; the card's BIN, bank, type, and whether it's commercial or foreign.

  • Your own priority

    You order amount and card rules in one list and priority decides; with no match the device's default bank is used.

  • Commission rates

    Rates per bank TID by card class and 1–36 instalments; routing scores cost with these rates.

  • Limits

    Daily, monthly and per-transaction limits; a bank row that's full drops out for that payment.

  • Health and circuit breaker

    Bank-caused failures in the last 24 hours lower the ranking; after consecutive failures a row closes briefly and returns through trial traffic.

  • Two AIs

    Autonomous AI learns cautiously, Adaptive AI more widely; or balanced, cost first, approval first, fastest and custom weights.

How it works
  1. 01

    Add banks

    Enter each bank's TID, commission rates and limits.

  2. 02

    Rules and strategy

    Order the rules, pick a strategy, try it in the simulator and publish.

  3. 03

    At payment time

    The engine picks the bank and writes it into the device command; if the engine doesn't answer, the default bank is used.

  4. 04

    Decide and learn

    Every decision is logged with its outcome; health and the model update from real results.

Connect your stores
to one API.

Let's plan your devices, banks and end-of-day setup together with our team.

For enterprise inquiries: [email protected]

Bank routing and AI — In-store POS | Lunixi