Skip to main content
Back to challenges
Challenge

Gas Field Expansion Risk Monitor with Google ADK & Gemini 2.5 Flash

Unconventional gas expansion megaprojects demand real-time analysis of drilling logs and supply chain delay notices. Build a multimodal risk evaluation agent using Google ADK powered by Gemini 2.5 Flash to flag project bottleneck risks.

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 an Agent Development Kit pipeline with Gemini 2.5 Flash to analyze gas field technical reports.

Delivery guide

How work is evaluated

Evaluation

Tests agent risk detection accuracy and cost calculation from Jafurah expansion documents.

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

severity_match

Verifies that delay_weeks equals 6 and severity is categorized correctly.

Binary check

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

Dimension 2

risk_precision

Precision score of classified risk levels against gold labels. • 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

What you should walk away with

  • Build a domain agent using Google ADK with multimodal document processing

  • Leverage Gemini 2.5 Flash for rapid ingestion of large engineering documents

  • Integrate Vertex AI agent tools for automated risk classification

  • Formulate quantitative cost impact estimates based on schedule delay logs

Resources and assets

Reference links and supporting material

Dataset notes

Contractual reports, vendor logs, and engineering delivery schedules for Jafurah unconventional gas development phase four.

Challenge setup

Tests agent risk detection accuracy and cost calculation from Jafurah expansion documents.

Challenge input

JSON containing vendor_log string and component_category

Google ADK

Google's Agent Development Kit for building agent workflows.

Gemini 2.5 Flash

High-speed multimodal LLM engine for reasoning over large contexts.

Evaluated output

JSON with risk_severity (CRITICAL|HIGH|MEDIUM|LOW), delay_weeks, and financial_impact_usd

Evaluation checks
  • Verifies that delay_weeks equals 6 and severity is categorized correctly.
  • Precision score of classified risk levels against gold labels. • target: 0.92 • range: 0-1
Configured evaluation evidence
  • Benchmark: Energy-SupplyChain-Bench
  • Risk Precision target: 0.92
  • 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
  • Google Gemini · Required
  • Google ADK · Optional
  • Google · Optional

Evaluation contract

  • severity_match · Weight 1
  • risk_precision · Weight 1

Metrics

  • risk_precision · Target 0.92

Benchmarks

  • Energy-SupplyChain-Bench
  • energy

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing vendor_log string and component_category

    Output: JSON with risk_severity (CRITICAL|HIGH|MEDIUM|LOW), delay_weeks, and financial_impact_usd

    Execution harness: Python

Evaluation tools

  • Google ADK · agent_framework
  • Gemini 2.5 Flash · llm_engine

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Gas Field Expansion Risk Monitor with Google ADK & Gemini 2.5 Flash