Why our refund flow never asks the model for permission
Language models are good at conversation and bad at policy. Here is how we split the two, and what it did to our error rate.
Engineering
How the platform is built, run and kept fast.
Sub-second voice: where the 800 milliseconds go
People notice a pause on the phone at around a second. This is the latency budget we work to, stage by stage.
PII redaction before the prompt, not after
If personal data reaches a model, no amount of filtering afterwards takes it back. Where we redact, and why it has to be first.