┌─────────────────────────────────────────────────────────────┐
│  TIER 1: ANTICIPATORY SHIPPING (Long-term — weeks/months)  │
│  → ML predicts demand → Moves inventory close to customers │
│     BEFORE they place an order                             │
├─────────────────────────────────────────────────────────────┤
│  TIER 2: REGIONALIZATION (Medium-term — days/weeks)        │
│  → Partitions the fulfillment network into autonomous zones│
│  → Ensures 70-80% of orders are fulfilled intra-region     │
├─────────────────────────────────────────────────────────────┤
│  TIER 3: CONDOR (Short-term — hours)                       │
│  → Continuously re-optimizes the fulfillment plan within   │
│     a 5-6 hour window before pick-and-pack begins.         │
└─────────────────────────────────────────────────────────────┘

Traditional Model:
  Customer orders → Warehouse processes → Ships → Delivered (2-5 days)

Anticipatory Shipping:
  ML predicts: "Customers in Region X will buy 200 iPhone 16s in the next 3 days"
  → Amazon ships 200 iPhones from a central hub to local delivery hubs in Region X
  → Customer places order → The item is already locally staged → Delivered same-day!

The Anticipatory Flow:

1. ML Model: "Zip code 10001 (NY) has an 87% probability of ordering 200 cases of water in the next 3 days."

2. System: Ships 200 cases from the Midwest Central Hub → NY Local Hub.
   These packages DO NOT HAVE A SPECIFIC CUSTOMER ADDRESS YET → They are just labeled "Destination: NY Hub".

3. Customer A in NY orders 2 cases of water:
   → The system assigns an address to 2 of the pre-staged cases at the NY Hub.
   → Delivered in 2 hours!

4. If predictions are slightly off (e.g., 50 cases remain unsold):
   → Amazon might run a targeted flash sale for that zip code.
   → Or re-route them back to the central hub.

17:00 — Order A is placed in Zone 1
  → CONDOR Plan v1: Ship from WH Alpha, individual truck.

17:15 — Order B is placed in Zone 1 (near Order A)
  → CONDOR Plan v2: Consolidate A+B onto the same route → saves 1 truck trip.

17:30 — Order C is placed in Zone 2 (along the same route)
  → CONDOR Plan v3: Consolidate A+B+C → highly dense, efficient route.

17:45 — Order D is placed in Zone 9 (opposite direction)
  → CONDOR Plan v4: Route 1 (A+B+C) + Route 2 (D only).

→ Every 15 minutes, CONDOR re-evaluates the entire network to find better plans.
→ Deadline: When the window closes (e.g., 23:00), the warehouse executes the final optimized plan.

PCVRP:
  Maximize: Total "prize" (value of orders delivered on time)
  Minimize: Total transportation cost
  Subject to:
    - Capacity constraints (vehicle limits)
    - Time windows (delivery SLAs)
    - Fleet size limits
    - Network density bonuses: bundling orders in the same neighborhood significantly reduces cost-per-package.

Solving Techniques:
  1. Mathematical optimization (LP/MIP relaxation)
  2. Local search heuristics (2-opt, 3-opt swaps between routes)
  3. Iterative re-optimization (running the solver continuously as new data arrives)

Pre-Regionalization:
  Customer in NY → Order fulfilled from CA (3,000 miles) → 3-5 day delivery.

Post-Regionalization:
  Customer in NY → Order fulfilled from NJ or PA (100 miles) → Same/Next day delivery.

Results:
  - Average travel distance per package dropped by ~60%
  - Significant reduction in shipping costs
  - Delivery times dropped by 1-2 days
  - Massive reduction in carbon footprint

SKU "IPHONE-16-256GB":
  National Demand: 100,000/month

  Northeast Region (NY, NJ, PA):  25,000/mo → Stock 30,000
  West Region (CA, WA, OR):       20,000/mo → Stock 25,000
  South Region (TX, FL, GA):      18,000/mo → Stock 22,000
  ...

  Buffer: 22,000 kept at a central cross-dock facility for overflow/rebalancing.