CrewAI & Bland AI Agritech DPI Advisory Agent
Smallholder farmers require rapid AI advisory services integrated with government agritech public infrastructure data. You will build a multi-agent workflow that ingests soil health cards, market pricing, and telemetry to deliver voice-based agricultural guidance via automated phone workflows.
What you are building
The core problem, expected build, and operating context for this challenge.
Orchestrate specialized CrewAI agents paired with Bland AI phone interfaces to process agritech public infrastructure data and automate voice advisory dispatch.
How work is evaluated
Validates multi-agent advisory generation and automated voice script compilation for agricultural support.
Challenge datasets
Public datasets, plus the private datasets you have access to.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
Call Dispatch Success
Checks that the Bland AI call dispatch returns a valid call ID
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
Agronomic Accuracy
Evaluates correctness of fertilizer ratios • target: 0.98 • range: 0.9-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Design multi-role CrewAI orchestration for soil data analysis and market linkage
Integrate Bland AI REST APIs for automated telephone voice calls to farmers
Process government Soil Health Card datasets with localized crop recommendations
Ensure voice conversational outputs adhere to rural dialect and language constraints
Reference links and supporting material
Agricultural soil health test readings paired with current mandi market prices across major Indian districts.
Challenge setup
Validates multi-agent advisory generation and automated voice script compilation for agricultural support.
Challenge input
JSON with farmer ID, language, N-P-K soil values, and crop preference
CrewAI
Orchestrates complex agent roles and tasks.
Bland AI
Executes natural language AI phone calls.
Evaluated output
JSON object with recommendations and Bland AI call ID
- Checks that the Bland AI call dispatch returns a valid call ID
- Evaluates correctness of fertilizer ratios • target: 0.98 • range: 0.9-1
- Benchmark: AgriAgent-Eval
- Agronomic Accuracy target: 0.98
- 1 public reference case
- 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
- Bland AI · Required
- CrewAI · Optional
- Guidance · Optional
Evaluation contract
- Call Dispatch Success · Weight 1
- Agronomic Accuracy · Weight 1
Metrics
- Agronomic Accuracy · Target 0.98
Benchmarks
- AgriAgent-Eval
- general
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON with farmer ID, language, N-P-K soil values, and crop preference
Output: JSON object with recommendations and Bland AI call ID
Execution harness: Python
Evaluation tools
- CrewAI · Agent Framework
- Bland AI · Voice Dispatch
Recipe freeze marker
No recipe freeze time is recorded.