Skip to main content
Back to challenges
Challenge

Riyadh Expo Substation RAG Assistant with LlamaIndex & ElevenLabs

Electrical engineering teams on gigaproject sites require immediate access to substation engineering schemas and vendor specifications. Build a hybrid search RAG agent using LlamaIndex that answers technical queries and delivers audio briefings via ElevenLabs.

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.

Combine LlamaIndex document stores with voice synthesis to query high-voltage substation engineering documentation.

Delivery guide

How work is evaluated

Evaluation

Measures response context recall and correctness for substation technical questions.

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

fact_accuracy_check

Checks if '63kA' and '380kV' exist in generated answer text.

Binary check

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

Dimension 2

retrieval_relevance

Cosine similarity of retrieved context nodes against golden reference. • target: 0.9 • range: 0-1

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

  • Build a structured RAG pipeline with LlamaIndex over substation technical specifications

  • Implement dynamic metadata filtering for engineering drawings and voltage ratings

  • Integrate ElevenLabs text-to-speech API for field audio briefing generation

  • Optimize retrieval accuracy using hybrid dense-sparse index structures

Resources and assets

Reference links and supporting material

Dataset notes

Substation project documentation including technical manuals, single-line diagrams descriptions, and vendor specs for high-voltage installations.

Challenge setup

Measures response context recall and correctness for substation technical questions.

Challenge input

JSON with query string and project_code

LlamaIndex

Framework used for indexing and document retrieval.

ElevenLabs

Generates clear field audio updates from text answers.

Evaluated output

JSON with answer text and audio_stream_url

Evaluation checks
  • Checks if '63kA' and '380kV' exist in generated answer text.
  • Cosine similarity of retrieved context nodes against golden reference. • target: 0.9 • range: 0-1
Configured evaluation evidence
  • Benchmark: RAG-Engineering-Eval
  • Retrieval Relevance target: 0.9
  • 1 public reference case
Configured runtime
  • 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

Action Space
  • ElevenLabs · Required
  • Llama Index · Optional
  • LangWatch · Optional

Evaluation contract

  • fact_accuracy_check · Weight 1
  • retrieval_relevance · Weight 1

Metrics

  • retrieval_relevance · Target 0.9

Benchmarks

  • RAG-Engineering-Eval
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with query string and project_code

    Output: JSON with answer text and audio_stream_url

    Execution harness: Python

Evaluation tools

  • LlamaIndex · rag_framework
  • ElevenLabs · voice_synthesis

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Riyadh Expo Substation RAG Assistant with LlamaIndex & ElevenLabs