{"data":{"kind":"file","path":"README.md","version_id":"cix1dszr5ua1wz9ta4fyzaky","entry":{"name":"README.md","path":"README.md","is_directory":false,"size":1746,"modified_at":"2026-08-11T02:05:24.175000","content_hash":"284860f41cf124eb7ba77aab27b13f19142a1c1d3f81d2e8838c24a2ce25e1ee"},"entries":[],"content":"# Customer Order Fulfillment Simulator\n\nA verifiers environment that trains LLMs to handle real-world customer order fulfillment scenarios — from inventory allocation and warehouse picking to shipping optimization and exception resolution.\n\n## Overview\n\nThe model receives structured order requests with constraints (inventory levels, shipping deadlines, customer tiers, warehouse locations) and must produce correct, efficient fulfillment plans. Each scenario tests different aspects of logistics decision-making.\n\n## Task Types\n\n- **Single-order fulfillment**: Allocate inventory, select carrier, estimate delivery\n- **Multi-order batch optimization**: Group orders for efficient warehouse picking and shipping\n- **Backorder handling**: Decide split-ship vs wait-for-restock based on customer priority\n- **Exception resolution**: Handle address errors, damaged items, carrier failures, returns\n- **Priority escalation**: VIP customer orders with expedited requirements\n\n## Reward Functions\n\n| Function | Weight | Description |\n|----------|--------|-------------|\n| `correctness_reward` | 1.0 | Core fulfillment plan is logically correct |\n| `efficiency_reward` | 0.5 | Shipping cost and warehouse effort minimized |\n| `completeness_reward` | 0.3 | All required fields and edge cases addressed |\n| `constraint_satisfaction` | 0.8 | All hard constraints (inventory, deadlines, rules) met |\n| `customer_tier_bonus` | 0.2 | Appropriate service level for customer tier |\n\n## Usage\n\n```python\nimport verifiers as vf\nenv = vf.load_environment(\"order-fulfillment\")\n# or\nfrom order_fulfillment import load_environment\nenv = load_environment()\n```\n\n## Evaluation\n\n```bash\nprime eval run capa/order-fulfillment -m Qwen/Qwen3.5-0.8B\n```\n\n## License\n\nMIT\n","encoding":"utf-8","truncated":false,"total_bytes":1746},"status":null}