Skip to main content
Loading navigation
Back to challenges

Gas Field Expansion Contract Risk Analyzer using LlamaIndex and Claude 4.1 Opus

Complex gas processing contract expansions require analyzing thousands of engineering standards and change order clauses under tight schedules. Build an advanced LlamaIndex RAG pipeline powered by Claude 4.1 Opus to extract schedule risk factors from contractual documentation, identifying schedule delay drivers with a baseline recall score above 90%.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Develop an advanced contract risk extraction pipeline utilizing LlamaIndex property graphs and Claude 4.1 Opus for mega gas field expansion engineering contracts.

How work is evaluated

Evaluation

Evaluates contract risk extraction recall and faithfulness on Aramco Jafurah EPC clauses.

Datasets

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

How submissions are scored

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

risk_penalty_detection

Ensure penalty percentage and max cap are correctly extracted

Binary check

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

Dimension 2

risk_extraction_recall

Recall metric of identified risk factors compared to standard ground truth • target: 0.92 • 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

  • Build a hierarchical LlamaIndex VectorStoreIndex and PropertyGraphIndex over EPC contract specs

  • Configure Claude 4.1 Opus as the primary LLM engine for complex clause reasoning

  • Extract delay risk penalties, force majeure, and long-lead material clauses into dynamic risk trees

  • Optimize retrieval chunking strategies using sub-question query engines

Resources

Dataset notes

EPC contract clauses, schedule penalty specs, and change order templates for Aramco Jafurah gas expansion project.

Challenge setup

Evaluates contract risk extraction recall and faithfulness on Aramco Jafurah EPC clauses.

Challenge input

Raw text excerpt from EPC contract phase IV agreement.

LlamaIndex

Required data framework for advanced RAG indexing

Claude 4.1 Opus

Required LLM model for high-reasoning contract parsing

Evaluated output

JSON containing detected risk clauses, max delay penalty %, and mitigation recommendations.

Evaluation checks
  • Ensure penalty percentage and max cap are correctly extracted
  • Recall metric of identified risk factors compared to standard ground truth • target: 0.92 • range: 0-1
Configured evaluation evidence
  • Benchmark: Legal RAG Analysis Benchmark
  • Risk Extraction Recall target: 0.92
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Shows the first task, up to three tools, and up to three evidence items.

View technical recipe

Configured tools

Action Space
  • Llama Index · Required
  • Exa · Optional
Policy Serving
  • Claude 4.1 Opus · Optional

Evaluation contract

  • risk_penalty_detection · Weight 1
  • risk_extraction_recall · Weight 1

Metrics

  • risk_extraction_recall · Target 0.92

Benchmarks

  • Legal RAG Analysis Benchmark
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: Raw text excerpt from EPC contract phase IV agreement.

    Output: JSON containing detected risk clauses, max delay penalty %, and mitigation recommendations.

    Execution harness: Python

Evaluation tools

  • LlamaIndex · Agent Framework
  • Claude 4.1 Opus · Model Runtime

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Gas Field Expansion Contract Risk Analyzer using LlamaIndex and Claude 4.1 Opus