Python Developer — Fix & Optimize Slot Detection Logic for 3AM Window + All-Day Monitoring ( Barq)

Job ID: 40617581

Budget: $10 – $100 USD

Please read the project carefully before submitting your proposal.

Python Developer — Fix & Optimize Slot Detection Logic for 3AM Window + All-Day Monitoring ( Barq)

Overview
We have a fully deployed booking system (Barq ) at https://*****ev.dev. The system is live and functional. We need one specific behavioral fix to the slot detection and booking logic.

The Exact Required Behavior (What Must Work):
Phase 1 — Intensive Window: 03:00 to 03:15 Baghdad Time (Daily)
This is the primary booking window. The server releases slots at 03:00 AM.
Slots may become bookable at any second between 03:00 and 03:15, not necessarily exactly at 03:00.
During this 15-minute window the system must run a precise, high-frequency scan (every 1–3 seconds) calling get_slots + confirming bookability via check-availability.
Priority order for booking:
First: slots for tomorrow (the next calendar day).
Second: any other available days within the current month.
The instant a slot is confirmed bookable → fire the booking POST immediately for all armed transactions targeting that office.
Do NOT wait until 03:00 exactly — start the intensive scan slightly before (e.g. 02:58) and continue until 03:15.
Phase 2 — All-Day Background Monitoring: 03:15 onwards (All Day)
After the 3AM window closes, the system continues a gentle background scan throughout the day.
servers occasionally release additional slots for tomorrow or scattered days within the month at unpredictable times during the day.
Scan frequency during the day: gentle and rate-limited (e.g. every 30–60 seconds) to avoid triggering 422 or 429 or 403 from the server.
Booking confirmation logic is identical: get_slots → confirm bookable via check-availability → fire immediately if bookable.
This scan runs continuously until midnight, then resets for the next 3AM window.
What is Already Working (Do Not Touch):
FastAPI backend, Cloudflare tunnel, watchdog, Telegram alerts — all running.
Token harvester via Frida feeding App Check tokens every 5 minutes.
Server-side fire engine with rotating Iraqi proxies (iq.decodo.com:44000).
Existing daytime monitoring exists in the system but needs the above logic adjustments.
capture_open_slots.py is available as a reference for how slot probing works against the real API.
What the Developer Must Deliver:
Fix the detection logic in the existing open_detector.py / fire_engine.py to implement the two-phase behavior above precisely.
Verify the slot confirmation step — a slot returned by get_slots is only acted upon if check-availability confirms it is actually bookable right now (not just displayed).
Test against a real 3AM window or simulate with the existing capture_open_slots.py tool to confirm the logic catches slots within seconds of them becoming available.
What We Provide:
Full codebase access.
Decodo proxy credentials and harvester secrets already configured.
Admin panel at *********ev.dev (login: admin / BARQ-js5Xg*****).
Existing slot capture logs from prior 3AM windows for reference.
Two live offices: 9117dadb (1021 الايمن), 0b3c251b (1022 الايسر). and all offices in my country
Required Skills:
Senior Python (FastAPI, Asyncio).
Experience with REST API polling logic and rate-limit-safe scanning strategies.
Ability to work within an existing codebase without rebuilding.
Scope:
One focused fix to the detection and booking trigger logic.
Deliverable: confirmed correct behavior at the next 3AM window. If there's a way to improve the appointment scheduling and booking process faster and better, I'm also willing to discuss it. Please provide examples of similar or comparable work and excellent experience.