Skip to main content
Back to challenges
Challenge

Anti-Scam Root Cause Analysis Crew with CrewAI & Together AI

Preventing public scams requires fast cross-analysis of reported phishing signals and telecom patterns. Deploy a collaborative agent crew using CrewAI and Together AI inference to automatically ingest scam alerts, identify root causes, and produce public threat advisories.

Workflow AutomationHosted 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 an anti-scam intelligence agent crew using CrewAI and high-throughput Together AI models.

Delivery guide

How work is evaluated

Evaluation

Evaluates threat categorization speed, root cause accuracy, and structured alert output.

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

correct_category

Checks category accuracy

Binary check

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

Dimension 2

root_cause_accuracy

Precision of root cause vector detection • target: 0.95 • 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

  • Configure CrewAI roles, tasks, and agents for threat intelligence

  • Integrate Together AI inference endpoints for open-weight model acceleration

  • Generate actionable public sector scam mitigation advisories

Resources and assets

Reference links and supporting material

Dataset notes

Synthetic dataset of SMS, email, and social messaging scam reports.

Challenge setup

Evaluates threat categorization speed, root cause accuracy, and structured alert output.

Challenge input

JSON containing reported SMS body and URL

CrewAI

Multi-agent crew orchestration

Together AI

Fast open-weights model execution

Evaluated output

JSON containing category, primary_vector, target_institution, recommended_action

Evaluation checks
  • Checks category accuracy
  • Precision of root cause vector detection • target: 0.95 • range: 0-1
Configured evaluation evidence
  • Benchmark: ScamThreat-Bench
  • Root Cause Accuracy target: 0.95
  • 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
  • RAI · Optional
Policy Serving
  • Together AI · Optional
Training Infra
  • Together AI · Optional

Evaluation contract

  • correct_category · Weight 1
  • root_cause_accuracy · Weight 1

Metrics

  • root_cause_accuracy · Target 0.95

Benchmarks

  • ScamThreat-Bench
  • security

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing reported SMS body and URL

    Output: JSON containing category, primary_vector, target_institution, recommended_action

    Execution harness: Python

Evaluation tools

  • CrewAI · Agent Framework
  • Together AI · Inference & Model Runtime

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Anti-Scam Root Cause Analysis Crew with CrewAI & Together AI