Skip to main content
Back to challenges
Challenge

Precision Agritech Dispatcher using LangChain and Vercel AI SDK

As agritech startups like Verdant Impact and Marut Drones scale, the need for intelligent coordination between sensor data and autonomous drone dispatching becomes critical. In this challenge, you will build a Precision Agritech Dispatcher using LangChain and the Vercel AI SDK. The goal is to create a stateful agent (using LangGraph) that monitors satellite moisture data and IoT soil sensors. When anomalies (e.g., pest alerts or localized drought) are detected, the agent must automatically orchestrate a drone 'mission plan' by querying Marut Drones' API for available hardware and flight regulations. The interface will be built with the Vercel AI SDK to provide a real-time, streaming dashboard for farmers, showing the agent's multi-step chain of thought as it moves from 'Detection' to 'Dispatch.'

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.

Build an agritech mission-control agent using LangChain and Vercel AI SDK to automate drone dispatch based on real-time soil and crop health data.

Delivery guide

How work is evaluated

Evaluation

Evaluation is based on mission plan validity: Did the agent correctly identify the crop issue and dispatch a drone with sufficient battery and the correct payload?

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

Safety Check

Drone must not be dispatched if battery < 30%.

Binary check

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

Dimension 2

Decision Latency

Time from alert to dispatch command. • target: 2000 • range: 0-5000

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

  • Develop stateful agentic workflows using LangGraph for multi-step mission planning

  • Implement real-time streaming AI interfaces using Vercel AI SDK (AI/RSC)

  • Integrate IoT sensor data (mocked) into LangChain Toolkits for contextual reasoning

  • Master the 'Manager Agent' pattern where one LLM coordinates specialized tool-calling agents

Resources and assets

Reference links and supporting material

Dataset notes

Time-series soil moisture and pest data for 50 fields, paired with a registry of 20 drones with varying specs.

Challenge setup

Evaluation is based on mission plan validity: Did the agent correctly identify the crop issue and dispatch a drone with sufficient battery and the correct payload?

Challenge input

JSON with field_data (soil_moisture, pest_level), available_drones (list), and current_weather

LangChain

Standard for building tool-calling chains and memory.

Vercel AI SDK

Provides best-in-class support for streaming LLM outputs to web apps.

Gemini 2.5 Flash

Fast and low-cost for high-frequency sensor analysis.

Evaluated output

JSON with selected_drone_id, flight_coordinates, and mission_type

Evaluation checks
  • Drone must not be dispatched if battery < 30%.
  • Time from alert to dispatch command. • target: 2000 • range: 0-5000
Configured evaluation evidence
  • Benchmark: Agentic-Industrial-IoT
  • Decision Latency target: 2000
  • 1 public reference case
Configured runtime
  • JavaScript/TypeScript execution harness
  • javascript 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
  • LangChain · Required
  • Langchain · Optional
  • Zed · Optional
Orchestration
  • LangChain · Required
  • Langchain · Optional

Evaluation contract

  • Safety Check · Weight 1
  • Decision Latency · Weight 1

Metrics

  • Decision Latency · Target 2000

Benchmarks

  • Agentic-Industrial-IoT
  • robotics

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with field_data (soil_moisture, pest_level), available_drones (list), and current_weather

    Output: JSON with selected_drone_id, flight_coordinates, and mission_type

    Execution harness: JavaScript/TypeScript

Evaluation tools

  • LangChain · Agent Orchestration
  • Vercel AI SDK · UI Streaming
  • Gemini 2.5 Flash · Inference Model

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Precision Agritech Dispatcher using LangChain and Vercel AI SDK