HTTP Request Optimization for Booking Bot

I already have a working headless automation bot that interacts with a booking-site frontend, but the browser layer is slowing everything down. I now want to strip that overhead away and drive the process purely with well-crafted HTTP requests.

Here is what I need:

• Re-engineer the current login flow so that authentication happens through direct requests rather than a rendered page.
• Recreate all form-selection steps (dates, guests, options, etc.) via the site’s underlying endpoints, keeping session integrity intact.
• Streamline every remaining call the bot makes so the full booking workflow is covered—from login to confirmation—without launching a browser.

The bot must still capture and manipulate two data sets: user information and reservation details. Speed is the primary metric; if we can cut execution time dramatically while maintaining reliability, the job is a success. I will supply the existing codebase, sample credentials, and traffic captures from the current headless run to help you map the required endpoints.

Deliverables:
1. Refactored scripts or modules that replace the headless portion with HTTP requests.
2. Clear documentation of each endpoint, headers, and payload used.
3. A short README with setup steps and a command that runs the full flow end-to-end, showing a successful booking in under the target time.

Feel free to use whatever stack you’re most productive with—Python requests, Node + Axios, Go’s net/http, or similar—so long as you keep the code readable and modular.