Skip to main content
Back to challenges
Challenge

Build a Gulf Corridor Logistics Exception Agent with LangChain and LangGraph

Geopolitical events and trade corridor shifts require instant rerouting and logistics exception management across Gulf ports and pipelines. Build a stateful graph agent using LangChain and LangGraph to detect route bottlenecks and trigger automated supply chain bypasses with 99% graph state recovery.

Business OperationsHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Develop a stateful cyclic logistics workflow using LangGraph state graphs and LangChain tools.

Delivery guide

How work is evaluated

Evaluation

Evaluates LangGraph node transitions and fallback rerouting logic when maritime choke points close.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

bypass_avoidance_check

Ensures blocked node Strait_of_Hormuz is not present in chosen_route array

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

route_optimality

Ratio of shortest valid path duration vs calculated duration • target: 0.98 • range: 0-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Construct a stateful routing state graph using LangGraph StateGraph primitives

  • Integrate LangChain tool wrappers for maritime transit time and pipeline throughput APIs

  • Implement persistent memory checkpointers for human-in-the-loop exception overrides

  • Automate rerouting decisions for disrupted maritime choke points (Strait of Hormuz bypasses)

Resources and assets

Reference links and supporting material

Dataset notes

Network graph nodes representing major GCC maritime ports, crude oil pipelines, and transit latency vectors.

Challenge setup

Evaluates LangGraph node transitions and fallback rerouting logic when maritime choke points close.

Challenge input

JSON containing origin_port, destination, cargo_type, blocked_nodes

LangChain

Tool wrappers and model abstractions

LangGraph

Stateful cyclic agent graphs

Evaluated output

JSON containing chosen_route, total_transit_days, cost_delta_usd

Evaluation checks
  • Ensures blocked node Strait_of_Hormuz is not present in chosen_route array
  • Ratio of shortest valid path duration vs calculated duration • target: 0.98 • range: 0-1
Configured evaluation evidence
  • Benchmark: Graph State Routing Bench
  • Route Optimality target: 0.98
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • LangGraph · Required
  • Langchain · Optional
  • AutoGluon · Optional
Orchestration
  • LangGraph · Required
  • Langchain · Optional

Evaluation contract

  • bypass_avoidance_check · Weight 1
  • route_optimality · Weight 1

Metrics

  • route_optimality · Target 0.98

Benchmarks

  • Graph State Routing Bench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing origin_port, destination, cargo_type, blocked_nodes

    Output: JSON containing chosen_route, total_transit_days, cost_delta_usd

    Execution harness: Python

Evaluation tools

  • LangChain · Agent Framework
  • LangGraph · Graph State Orchestrator

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build a Gulf Corridor Logistics Exception Agent with LangChain and LangGraph