Skip to main content
Back to challenges
Challenge

Bilingual GCC Oil Trade Settlement & Compliance Crew with CrewAI

Oil trade execution under short-term mechanisms demands seamless translation, legal cross-checking, and tax regulatory verification across Arabic decrees and English contracts. Build an autonomous multi-role agent crew using CrewAI. The crew consists of three agents: Arabic Legal Translator Agent, SOMO Regulatory Auditor, and Trade Settlement Manager. The crew collaborates sequentially to translate incoming Arab contracts, assess compliance with the three-month window, and output a validated bilingual clearance certificate.

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.

Orchestrate a sequential multi-role crew for bilingual oil trade settlement using CrewAI.

Delivery guide

How work is evaluated

Evaluation

Evaluates crew task completion quality and translation compliance precision.

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

check_clearance_code

Verify clearance code is present and valid.

Binary check

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

Dimension 2

translation_semantic_similarity

Semantic similarity of translated summary to gold standard. • target: 0.94 • 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

  • Define CrewAI agents, tasks, and process workflows.

  • Implement sequential task execution with state passing between agents.

  • Process Arabic and English legal oil trade documentation.

Resources and assets

Reference links and supporting material

Dataset notes

20 bilingual Arabic-English petroleum contract snippets and official SOMO decree texts.

Challenge setup

Evaluates crew task completion quality and translation compliance precision.

Challenge input

JSON with arabic_contract_text and contract_val_usd

CrewAI

Structured role-playing framework for sequential business workflows.

Evaluated output

JSON with English_summary, compliance_pass, clearance_code

Evaluation checks
  • Verify clearance code is present and valid.
  • Semantic similarity of translated summary to gold standard. • target: 0.94 • range: 0-1
Configured evaluation evidence
  • Benchmark: CrewAI-Bilingual-Bench
  • Translation Semantic Similarity target: 0.94
  • 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
  • CrewAI · Required
  • crewAI · Optional
Policy Serving
  • OpenRouter · Optional
Orchestration
  • OpenRouter · Optional

Evaluation contract

  • check_clearance_code · Weight 1
  • translation_semantic_similarity · Weight 1

Metrics

  • translation_semantic_similarity · Target 0.94

Benchmarks

  • CrewAI-Bilingual-Bench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with arabic_contract_text and contract_val_usd

    Output: JSON with English_summary, compliance_pass, clearance_code

    Execution harness: Python

Evaluation tools

  • CrewAI · Multi-Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Bilingual GCC Oil Trade Settlement & Compliance Crew with CrewAI