Priority-aware resource scheduling system for high-stakes operational environments. Assigns 180 meeting requests to 20 conference rooms and 150 personnel across 5 departments using a constrained greedy algorithm with conflict detection. Runs entirely in the browser — no backend required.
Live demo: akbknight.github.io/scheduling-optimizer
Resource scheduling under hard constraints is one of the hardest operational planning problems — especially in high-stakes environments where double-booking a room or scheduling conflicting personnel can have real consequences. This tool demonstrates:
- Priority queuing: High-priority meetings are scheduled first and get first pick of available rooms
- Multi-constraint satisfaction: Room capacity, security clearance levels, and attendee availability must all be satisfied simultaneously
- Conflict detection: Before any assignment, all three constraints are checked — room availability, personnel busy schedules, and clearance authorization
- Utilization optimization: After the initial pass, a sweep fills remaining gaps with lower-priority meetings
Synthetic data — no real personnel or operational data is used:
- 20 conference rooms (6–30 capacity, clearance levels 1–5)
- 150 personnel across 5 departments: State, Consular, Admin, Security, Protocol
- 180 meeting requests with priority scores (1–5), duration (30 min – 2 hrs), attendee lists, and clearance requirements
- Scheduling window: 8:00 AM – 6:00 PM (20 × 30-minute slots)
1. Score all meetings by: priority × clearance_weight × urgency
2. Sort by score descending (highest priority scheduled first)
3. For each meeting:
a. Find earliest available room satisfying capacity + clearance
b. Check all required attendees are free for the full duration
c. If valid: assign room + mark attendees busy
d. If no valid slot found: mark as queued for rescheduling
4. Secondary sweep: fill room gaps with queued lower-priority meetings
Typical result: ~85% of meetings successfully scheduled, average room utilization ~65%, all conflicts resolved before assignment.
This system mirrors the architecture of scheduling infrastructure I built for the U.S. Department of State in New Delhi — serving 1,200+ diplomatic and support personnel across consular operations, administrative workflows, and security protocols. The real system ran as an Azure-backed web application with Power Apps UI; this demo reconstructs the constraint logic as a client-side simulation.
- Vanilla JavaScript (no framework, no build step)
- Chart.js 4 (utilization bar chart)
- Seeded RNG (mulberry32) — reproducible synthetic data
- Gantt visualization in pure CSS/DOM
- Operations research: Greedy constrained scheduling, priority queue
- Algorithm design: Multi-constraint satisfaction with conflict detection
- Data visualization: Gantt chart, utilization metrics, meeting table
- Systems thinking: Translating a real enterprise scheduling problem into a clean algorithmic model
Akshay Kumar
linkedin.com/in/akshaykumardl