Skip to main content
Loading navigation
Back to challenges

Build a BESS Dispatch Optimizer Agent with Mastra AI & Hume AI

Grid stability during peak energy shifts in Saudi Arabia requires real-time battery energy storage optimization. Build an interactive BESS dispatch agent using Mastra AI and Hume AI to voice-query grid status and automate dispatch scheduling across 1.16bn BESS assets.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Implement a Mastra AI workflow combined with Hume AI empathetic audio interface to deliver natural language grid management and BESS charge orchestration.

How work is evaluated

Evaluation

Tests the Mastra workflow logic for state-of-charge optimization and audio response generation.

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

valid_action_check

Check if action is CHARGE, DISCHARGE, or IDLE

Binary check

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

Dimension 2

efficiency_gain

Percentage efficiency optimization • target: 0.92 • range: 0.7-1

Binary check

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

Learning goals

  • Develop Mastra AI dynamic workflows for battery state-of-charge calculation

  • Integrate Hume AI voice interface for dispatcher interactive queries

  • Simulate high-capacity battery charge-discharge cycles against grid demand spikes

  • Output JSON execution telemetry for SCADA control integration

Resources

Dataset notes

Telemetry logs from grid-scale energy storage projects including state of charge, temperature, and hourly grid load profiles.

Challenge setup

Tests the Mastra workflow logic for state-of-charge optimization and audio response generation.

Challenge input

JSON grid telemetry

Mastra AI

TypeScript-native agent workflow engine

Hume AI

Enables natural voice interaction for grid dispatchers

Evaluated output

JSON dispatch order

Evaluation checks
  • Check if action is CHARGE, DISCHARGE, or IDLE
  • Percentage efficiency optimization • target: 0.92 • range: 0.7-1
Configured evaluation evidence
  • Benchmark: GridStorageBench
  • Efficiency Gain target: 0.92
  • 1 public reference case
Configured runtime
  • JavaScript execution harness
  • javascript 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
  • Hume AI · Required
  • Mastra AI · Optional
  • Twin · Optional
Orchestration
  • Mastra AI · Optional

Evaluation contract

  • valid_action_check · Weight 1
  • efficiency_gain · Weight 1

Metrics

  • efficiency_gain · Target 0.92

Benchmarks

  • GridStorageBench
  • energy

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON grid telemetry

    Output: JSON dispatch order

    Execution harness: JavaScript

Evaluation tools

  • Mastra AI · Agent Framework
  • Hume AI · Voice AI Platform

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build a BESS Dispatch Optimizer Agent with Mastra AI & Hume AI