BESS Grid Integration Multi-Agent System with CrewAI & Langfuse
Saudi Arabia's $1.16bn BESS investment requires dynamic dispatch orchestration to balance renewable intermittent supply against grid load demands. Build a CrewAI agentic workflow integrated with Langfuse monitoring to automatically parse grid telemetry, predict storage needs, and execute optimal dispatch commands. Achieve at least 95% dispatch accuracy while keeping latency under two seconds.
What you are building
The core problem, expected build, and operating context for this challenge.
Orchestrate multi-agent BESS energy balancing using CrewAI role-playing agents monitored in real-time via Langfuse.
How work is evaluated
Evaluates CrewAI agent dispatch actions against simulated BESS grid telemetry and verifies trace logging in Langfuse.
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.
Safety Limits Check
Ensures battery discharge does not breach minimum SoC threshold (10%).
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
Dispatch Decision Accuracy
Percentage of optimal power actions taken based on telemetry rules. • target: 0.95 • range: 0-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Deploy CrewAI agents for BESS dispatch and grid demand forecasting
Trace LLM reasoning and latency using Langfuse observability
Implement robust error handling for real-time grid telemetry processing
Validate battery state-of-charge limits and state constraints programmatically
Reference links and supporting material
Simulated 15-minute resolution power telemetry feed for Saudi electricity grid nodes connected to 1.16bn BESS infrastructure.
Challenge setup
Evaluates CrewAI agent dispatch actions against simulated BESS grid telemetry and verifies trace logging in Langfuse.
Challenge input
JSON telemetry payload containing grid_freq_hz, solar_gen_mw, demand_mw, current_soc_pct
CrewAI
Mandatory role-playing agent orchestration framework.
Langfuse
Required tracking framework for agent decisions.
Evaluated output
JSON response with action (charge/discharge/idle), rate_mw, and reasoning string
- Ensures battery discharge does not breach minimum SoC threshold (10%).
- Percentage of optimal power actions taken based on telemetry rules. • target: 0.95 • range: 0-1
- Benchmark: EnergyAgent-Bench
- Dispatch Decision Accuracy target: 0.95
- 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
- CrewAI · Optional
- ElevenLabs · Optional
- Langfuse · Required
- Langfuse · Required
Evaluation contract
- Safety Limits Check · Weight 1
- Dispatch Decision Accuracy · Weight 1
Metrics
- Dispatch Decision Accuracy · Target 0.95
Benchmarks
- EnergyAgent-Bench
- energy
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON telemetry payload containing grid_freq_hz, solar_gen_mw, demand_mw, current_soc_pct
Output: JSON response with action (charge/discharge/idle), rate_mw, and reasoning string
Execution harness: Python
Evaluation tools
- CrewAI · Agent Framework
- Langfuse · LLM Observability
Recipe freeze marker
No recipe freeze time is recorded.