Office Conversion Feasibility Engine using LangChain and Relicx Workflow Testing
Commercial office-to-residential conversions often encounter budget overruns due to undisclosed structural load limits and egress bottlenecks. Build a LangChain structural audit agent integrated with Relicx test automation to parse CAD metadata and verify building feasibility with 90% accuracy.
What you are building
The core problem, expected build, and operating context for this challenge.
Implement a stateful LangChain workflow that evaluates structural architectural blueprints against residential building code constraints while validating user interactions via Relicx.
How work is evaluated
Evaluates the precision of structural feasibility scoring and the pass rate of Relicx workflow assertions.
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.
Egress Compliance Accuracy
Verifies that egress non-compliance is correctly detected.
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
feasibility_score_mae
Mean Absolute Error between predicted feasibility score and expert ground truth • target: 5 • range: 0-100
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 a LangGraph agentic pipeline to extract structural constraints from building blueprints.
Integrate Relicx workflow assertions to test user-facing conversion feasibility reports.
Calculate floor plate depth and HVAC vertical shaft conversion cost penalties.
Achieve automated validation of residential egress compliance against regional building codes.
Reference links and supporting material
Synthetic dataset of 150 commercial office building floor plans with annotated perimeter depths, shaft counts, window retrofitting flags, and actual conversion cost metrics.
How this agent runs
Evaluates the precision of structural feasibility scoring and the pass rate of Relicx workflow assertions.
Challenge input
JSON containing floor plate depth, core distance, and window perimeter
LangChain
Required framework for building stateful multi-step agent reasoning chains.
Relicx
Required tool for testing end-to-end agent workflow execution and dashboard renders.
Evaluated output
JSON with feasibility_score (0-100), estimated_cost_sqft, and egress_compliant (boolean)
- Verifies that egress non-compliance is correctly detected.
- Mean Absolute Error between predicted feasibility score and expert ground truth • target: 5 • range: 0-100
- Benchmark: CRE Structural Feasibility Benchmark
- Feasibility Score Mae target: 5
- 1 public reference case
- Python execution harness
- Python sandbox
View technical recipe
Configured tools
- Relicx · Required
- Langchain · Optional
- All Hands AI · Optional
- Langchain · Optional
Evaluation contract
- Egress Compliance Accuracy · Weight 1
- feasibility_score_mae · Weight 1
Recipe state
This is a preview. The configuration can change before the evaluation recipe is locked.