The master schedule produced a total of 1,543 scheduled interviews and 1,420 diagnosed unscheduled interviews.
The API returned the schedule generation response in 164ms.
Both replanning and defense simulation responses were delivered in 185ms.
I built the ProPlacement repository on GitHub under my username @nanthitha25. It is a Python project started in August 2026 and last updated the same month.
proplacement.repo_name = ProPlacement
proplacement.owner = @nanthitha25
proplacement.language = Python
proplacement.project_start = 2026-08
proplacement.last_update = 2026-08
The system is a placement‑week scheduler & live‑defense replanner for university placement drives, promising Zero Silent Failures, enforcing the Minimal Churn Principle, and being Defense‑Ready for compound live disruptions.
proplacement.claims = Zero Silent Failures, Minimal Churn Principle, Defense‑Ready for compound live disruptions
I chose a decoupled architecture: a FastAPI backend hosts the optimization engine, while a React 18 + Tailwind CSS frontend provides a real‑time cockpit, separating computation from UI for high performance.
proplacement.architecture = decoupled FastAPI backend + React 18/Tailwind CSS frontend
proplacement.scheduler_collision_state = O(1) bitmask
proplacement.frozen_time_buffer = 30‑minute
proplacement.disruption_pipeline_stages = 4
The data generator (Pillar 1) creates a realistic campus dataset using a truncated normal distribution for CGPA and a Zipf‑Pareto distribution for shortlist skew.
proplacement.dataset = 800 Students, 35 Companies, 20 Rooms
proplacement.timeline_matrix = 20 Rooms × 4 Days
proplacement.example_disruption = TCS 3h Late + Panel Drop + 15 Withdrawals