Build a Multi-Site Battery Storage Portfolio Manager with LlamaIndex
Managing performance across three BESS deals worth $867m requires unified multi-site telemetry aggregation. Build an agentic RAG solution using LlamaIndex workflows to track battery degradation, query maintenance manuals, and optimize warranty coverage.
What you are building
The core problem, expected build, and operating context for this challenge.
Utilize LlamaIndex Workflows and agentic tools to monitor multi-site BESS telemetry and match fault logs against warranty terms.
How work is evaluated
Evaluates fault code diagnosis accuracy and warranty claim generation correctness.
Shared data for this challenge
Review public datasets and any private uploads tied to your build.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
warranty_coverage_check
Verify warranty status is correctly identified as covered
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
diagnosis_accuracy
Accuracy of fault code classification against manuals • 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
Construct LlamaIndex Event-driven Workflows for multi-site telemetry ingestion
Index technical warranty manuals and cell degradation lookup tables
Automate fault diagnosis when battery cell voltage variance exceeds tolerance limits
Generate automated warranty claim submissions for regional BESS suppliers
Reference links and supporting material
SCADA fault logs, battery cell SOH degradation curves, and manufacturer warranty contracts across 3 Saudi BESS sites.
How this agent runs
Evaluates fault code diagnosis accuracy and warranty claim generation correctness.
Challenge input
JSON BMS telemetry log
LlamaIndex
Offers event-driven Workflow architecture for multi-document RAG
Evaluated output
JSON warranty assessment
- Verify warranty status is correctly identified as covered
- Accuracy of fault code classification against manuals • target: 0.98 • range: 0.9-1
- Benchmark: BMSFaultBench
- Diagnosis Accuracy target: 0.98
- 1 public reference case
- Python execution harness
- Python sandbox (unavailable on Versalist)
View technical recipe
Configured tools
- Llama Index · Required
- Ray Tune · Optional
- Hume AI · Optional
Evaluation contract
- warranty_coverage_check · Weight 1
- diagnosis_accuracy · Weight 1
Recipe state
This is a preview. The configuration can change before the evaluation recipe is locked.