Where a request goes
A call tohttps://api.sqwish.ai or a visit to https://console.sqwish.ai uses TLS. Cloudflare’s network is in front of the API. Its tunnel carries the request, encrypted, to the API, and it does not store the body.
The API sends the request to a GPU that can serve the model you asked for. That hop is on the private network, and the GPUs accept connections only over that network. The GPU reads the context and the questions. It returns a probability for each allowed answer, and sends that score back to the API.
The models are decision models in three sizes, from kettei-fat to kettei-core.
If the model you asked for can’t answer, another ready size answers by default, and the answer’s fallback says so (fallback). Send "fallback": "none" to get the error instead, and read its error.retryable (errors). The answer’s model field is the size that scored it.
What happens to the text
You own the context, the questions and the files you upload. You allow Sqwish Labs to store and process them to run the service for you. They are not used to train the models offered to anyone else. A model fine-tuned on your data can be called only by your organisation. The API answers and discards the content of a request. An answer is kept for retries only when the request carries an idempotency key, which the Python and TypeScript clients send with every decision (retries). Nothing from a call is kept for training.What is kept, and where
A successful answer to a request sent with an
Idempotency-Key is also kept for 24 hours. A retry with the same key and body gets it back without running or charging again. After 24 hours the key won’t run the decision again (idempotency keys). An organisation with zero data retention for inference keeps none of these answers, for the requests it sends once the switch has taken effect.
