CADi · Payment decisions and customer support
Find the hidden cost of payment decisions.
Investigate false declines, fraud, routing and retry, alongside the support given to customers in difficulty or vulnerability.
CADi Pay
See how CADi Pay governs payment decisions at machine speed while keeping evidence, policy and accountability attached.
CADi Pay is available for deployment. Historical analysis and SideCar estimates are analytical, not guaranteed outcomes. HotPath authority is enabled only after client-specific validation and governance.
A good customer wrongly declined may leave no complaint and no fraud report. The lost sale is visible only as a transaction that did not happen. Tightening a control can improve one metric while creating costs elsewhere.
The same problem reaches customer support. A payment decision can look efficient while leaving someone in difficulty without the help they need. CADi Pay investigates both the financial consequences and the customer outcomes.
Begin with the question and your decision history. A live evidence feed or live decision authority can follow when the deployment evidence and controls support it.
01 · Understand
Share the relevant payment and support history. We examine the decisions, estimate alternative costs where supported, and return findings, recommendations and the evidence behind them. This stage needs no connection to live systems.
02 · Observe
A CADi SideCar feed absorbs your live decision data asynchronously: near real time, read-only, zero touch on the payment path. Decisions stream through the engine as they flow, each one explained, priced and recorded, and the audit trail builds itself instead of arriving in batches.
03 · Govern
A later deployment stage: your system consults CADi before an agreed decision. Live authority is enabled only after client-specific validation, controls and accountability are in place.
Authorise is the transaction moment: approve, decline, step up authentication, route or retry. CADi Pay examines the evidence behind those choices and estimates the consequences of alternatives where the model and data support them.
Treat is the customer moment: what support is appropriate when someone faces payment difficulty or vulnerability? We examine the decision, its reasoning and the customer outcome alongside your existing payment and support systems.
Identify potential false-decline losses and compare alternatives, with financial estimates tied to their assumptions and evidence.
Examine support decisions and differences across customer groups, including people experiencing difficulty or vulnerability.
Compare routing, retry and treatment choices and their estimated costs in a decision-level ledger.
Keep the evidence, policy, rationale and tamper-evident trace together so a payment or support decision can be reviewed.
PSPs, gateways and the acquirers behind them, where a one-to-two-point swing in authorisation rate is a board-level number and false-decline pain is real but unmeasured.
Regulated lenders and customer-operations teams who must evidence good, proportionate outcomes for customers in payment difficulty and vulnerability.
Did fraud fall because protection improved, or because more good customers were turned away? Did a support change help customers, or simply move the problem to another team?
CADi Pay investigates those questions using explicit causal models and the relevant decision history. The findings show what the evidence supports, what remains uncertain and which changes are worth testing.
CADi is in production - its engines are live in the cloud - with nine UK patent applications pending. CADi Pay aligns with the frameworks it touches: PSR reimbursement rules, FCA Consumer Duty, SCA and exemption regimes. It never claims to "guarantee compliance" or ensure a regulatory outcome.
Causal analysis runs on a declared model of the decision environment. Its conclusions depend on that model, the available evidence and explicit assumptions; estimates are analytical approximations. The engine has been tested against injected, synthetic and real ingested data. Live-outcome validation is completed against each client's own outcomes during deployment.
Bring us a problem, a suspected opportunity or a result that does not add up. We can explore how CADi Pay could help, starting with your question and the evidence available.
Get in touch