Skip to main content
Loading navigation
Back to challenges

Office-to-Residential Structural Conversion Agent with LlamaIndex & Mastra AI

Converting obsolete office buildings into housing often fails due to structural constraints such as deep floorplates and plumbing stack limits. Build a specialized spatial indexing and document retrieval system using LlamaIndex and Mastra AI to evaluate architectural blueprints against residential conversion codes. Deliver conversion feasibility scores in seconds.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Construct an automated structural feasibility parser for commercial adaptive reuse using LlamaIndex RAG and Mastra AI workflow nodes.

How work is evaluated

Evaluation

Evaluates structural feasibility assessments generated from architectural floorplate attributes.

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

Challenge Extraction

Ensures deep floorplates correctly trigger daylighting warnings

Binary check

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

Dimension 2

Feasibility Score Variance

Absolute error compared to professional engineering benchmarks • target: 3 • range: 0-10

Binary check

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

Learning goals

  • Parse multi-page structural blueprints and BIM CAD metadata using LlamaIndex data connectors

  • Formulate spatial evaluation graphs in Mastra AI to test plumbing and daylighting compliance

  • Extract structural load bearing wall positions to flag column grid restrictions

  • Generate cost-per-square-foot feasibility estimates for adaptive reuse

Resources

Dataset notes

BIM model metadata and architectural specifications of 1970s commercial office buildings.

Challenge setup

Evaluates structural feasibility assessments generated from architectural floorplate attributes.

Challenge input

JSON with floorplate_depth_ft, core_distance_ft, window_ratio

LlamaIndex

Indexes architectural specifications and building code documents

Mastra AI

Executes evaluation workflow nodes in TypeScript/Python

Evaluated output

JSON with feasibility_score (0-100), structural_challenges

Evaluation checks
  • Ensures deep floorplates correctly trigger daylighting warnings
  • Absolute error compared to professional engineering benchmarks • target: 3 • range: 0-10
Configured evaluation evidence
  • Benchmark: StructuralFeasibilityBench
  • Feasibility Score Variance target: 3
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python 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
  • Mastra AI · Required
  • Llama Index · Optional
  • Zed · Optional
Orchestration
  • Mastra AI · Required

Evaluation contract

  • Challenge Extraction · Weight 1
  • Feasibility Score Variance · Weight 1

Metrics

  • Feasibility Score Variance · Target 3

Benchmarks

  • StructuralFeasibilityBench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with floorplate_depth_ft, core_distance_ft, window_ratio

    Output: JSON with feasibility_score (0-100), structural_challenges

    Execution harness: Python

Evaluation tools

  • LlamaIndex · Data Indexing & RAG
  • Mastra AI · Agent Workflow Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Office-to-Residential Structural Conversion Agent with LlamaIndex & Mastra AI