Skip to main content
Back to challenges
Challenge

Build a Chemical & Oilfield Project Safety Agent with AutoGen

Managing high-risk chemical formulation in Oman's Nizwa Industrial City ($10.4m) requires strict safety inspection protocols. Build a multi-agent system using Microsoft AutoGen to review chemical safety data sheets (MSDS), monitor hazardous material storage, and enforce compliance.

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.

Utilize Microsoft AutoGen agent conversations to audit oilfield chemical storage layouts and hazard prevention standards.

Delivery guide

How work is evaluated

Evaluation

Evaluates detection of chemical storage incompatibilities and safety buffer distance verification.

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

hazard_detection_check

Ensure non-compliant chemical pair is correctly flagged

Binary check

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

Dimension 2

hazard_classification_precision

Precision of GHS hazard code classification • target: 0.99 • range: 0.92-1

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

  • Configure AutoGen Safety Auditor and Chemical Process Engineer agents

  • Parse Material Safety Data Sheets (MSDS) for oilfield chemical formulations

  • Verify minimum separation distances between volatile reactive chemical tanks

  • Generate automated OSHA and Omani industrial safety compliance certificates

Resources and assets

Reference links and supporting material

Dataset notes

MSDS sheets, hazardous substance registries, and facility site layouts for Nizwa Industrial City chemical units.

Challenge setup

Evaluates detection of chemical storage incompatibilities and safety buffer distance verification.

Challenge input

JSON chemical storage config

AutoGen

Microsoft framework for orchestrating conversational safety inspection teams

Evaluated output

JSON audit result

Evaluation checks
  • Ensure non-compliant chemical pair is correctly flagged
  • Precision of GHS hazard code classification • target: 0.99 • range: 0.92-1
Configured evaluation evidence
  • Benchmark: HazMatSafetyBench
  • Hazard Classification Precision target: 0.99
  • 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
  • AutoGen · Required
  • Microsoft AutoGen · Optional
Observation
  • Langfuse · Optional
Reward / Eval
  • Langfuse · Optional

Evaluation contract

  • hazard_detection_check · Weight 1
  • hazard_classification_precision · Weight 1

Metrics

  • hazard_classification_precision · Target 0.99

Benchmarks

  • HazMatSafetyBench
  • security

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON chemical storage config

    Output: JSON audit result

    Execution harness: Python

Evaluation tools

  • AutoGen · Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build a Chemical & Oilfield Project Safety Agent with AutoGen