Skip to main content
Loading navigation
Back to challenges

Build a CrewAI Agritech Advisory & Subsidy Verification Crew

Smallholder farmers require real-time agricultural advisories paired with digital subsidy validation across government agritech portals. Build a multi-agent CrewAI system that verifies soil data, assesses crop health, and cross-references farmer welfare schemes with 90% recommendation accuracy.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Construct a collaborative multi-agent CrewAI system combining agronomist and subsidy compliance specialist roles for smallholder farmers.

How work is evaluated

Evaluation

Evaluates multi-agent task execution and crop recommendation correctness.

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

recommendation_check

Verifies valid advisory output

Binary check

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

Dimension 2

accuracy_score

Agronomist recommendation correctness • 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

  • Define specialized Agent roles in CrewAI for Agronomist, Scheme Auditor, and Soil Specialist

  • Orchestrate sequential and hierarchical Tasks in CrewAI workflows

  • Extract land parcel data and verify crop subsidy eligibility requirements

  • Output structured agritech advisory cards for farmer mobile applications

Resources

Dataset notes

Synthetic dataset containing regional soil sensor readings and state agricultural subsidy rules.

Challenge setup

Evaluates multi-agent task execution and crop recommendation correctness.

Challenge input

JSON soil test data and land location

CrewAI

Role-playing agent orchestration framework

Evaluated output

JSON advisory recommendation with crop_choice and subsidy_eligible

Evaluation checks
  • Verifies valid advisory output
  • Agronomist recommendation correctness • target: 0.92 • range: 0-1
Configured evaluation evidence
  • Benchmark: Agritech Multi-Agent Benchmark
  • Accuracy Score 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
  • CrewAI · Required
  • crewAI · Optional
  • Kore.ai · Optional

Evaluation contract

  • recommendation_check · Weight 1
  • accuracy_score · Weight 1

Metrics

  • accuracy_score · Target 0.92

Benchmarks

  • Agritech Multi-Agent Benchmark
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON soil test data and land location

    Output: JSON advisory recommendation with crop_choice and subsidy_eligible

    Execution harness: Python

Evaluation tools

  • CrewAI · Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build a CrewAI Agritech Advisory & Subsidy Verification Crew