Build Urban Site Selection and Zoning Compliance Agent using LlamaIndex
Selecting optimal sites for urban regeneration projects—such as King's Cross—requires synthesizing transit accessibility, historic land use, and local zoning parameters. Developers will build a site selection agent using LlamaIndex. The agent indexes municipal urban plans, transit density maps, and environmental hazard filings. It scores candidate sites based on redevelopment feasibility, transit proximity, and zoning compliance, automating initial feasibility studies for master plan developers.
What you are building
The core problem, expected build, and operating context for this challenge.
Develop an urban site selection agent using LlamaIndex to score development feasibility across spatial and regulatory datasets.
How work is evaluated
Evaluates site feasibility scoring accuracy and retrieval precision across zoning documents.
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.
Zoning Compliance Check
Verifies that proposed FAR <= allowed FAR yields zoning_compliant = True
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
Feasibility Score Precision
Fidelity of feasibility calculation • 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
Index municipal zoning codes and spatial transit datasets into LlamaIndex vector stores
Build custom retrieval tools for filtering parcels by zoning allowance and FAR (Floor Area Ratio)
Calculate composite Feasibility Scores for candidate urban development sites
Generate automated site selection executive summaries
Reference links and supporting material
Dataset of 50 urban parcel records with zoning codes, FAR constraints, and transit connectivity metrics.
Challenge setup
Evaluates site feasibility scoring accuracy and retrieval precision across zoning documents.
Challenge input
JSON with parcel_id, allowed_far, proposed_far, transit_score, environmental_risk
LlamaIndex
Required framework for document retrieval.
Evaluated output
JSON with feasibility_score, zoning_compliant (boolean), primary_advantage
- Verifies that proposed FAR <= allowed FAR yields zoning_compliant = True
- Fidelity of feasibility calculation • target: 0.95 • range: 0-1
- Benchmark: SiteSelectBench
- Feasibility Score Precision 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
- Llama Index · Required
- Weights & Biases · Optional
- Hume AI · Optional
Evaluation contract
- Zoning Compliance Check · Weight 1
- Feasibility Score Precision · Weight 1
Metrics
- Feasibility Score Precision · Target 0.95
Benchmarks
- SiteSelectBench
- general
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON with parcel_id, allowed_far, proposed_far, transit_score, environmental_risk
Output: JSON with feasibility_score, zoning_compliant (boolean), primary_advantage
Execution harness: Python
Evaluation tools
- LlamaIndex · Agent Framework
Recipe freeze marker
No recipe freeze time is recorded.