Build Rate-Sensitivity Home Sales Predictor with Vercel AI SDK & Qwen 3
High interest rates create consumer friction, leading homebuilders to experience high foot-traffic but lagging sales conversions. Developers need predictive real-time interactive UI dashboards that dynamically model home buyer rate sensitivity and suggest optimized builder incentives. In this challenge, developers will use Vercel's AI SDK combined with Qwen 3 (via API) to construct a real-time web interface. The agent processes regional interest rate feeds, buyer debt profiles, and builder inventory, automatically outputting dynamic rate concessions, price reductions, or design studio credit strategies to close sales.
What you are building
The core problem, expected build, and operating context for this challenge.
Construct an interactive home sales conversion engine using Vercel AI SDK and Qwen 3 to simulate buyer rate sensitivity.
How work is evaluated
Evaluates buyer conversion likelihood score and optimal incentive package determination under various mortgage rate inputs.
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.
Budget Constraint Verification
Verifies calculated payment is strictly within buyer_max_monthly_budget
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
Conversion Predictor AUC
Accuracy of buyer conversion prediction based on budget fit • target: 0.92 • 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
Implement streaming UI tool calling using Vercel AI SDK (`ai/react`)
Utilize Qwen 3 model runtime for dynamic multi-scenario financial reasoning
Build real-time buyer affordability simulation logic for sales agents during client interactions
Deploy interactive pricing recommendation components for real estate CRM dashboards
Reference links and supporting material
Dataset of 150 buyer financial profiles and home sale transaction logs under fluctuating 2024-2025 interest rates.
Challenge setup
Evaluates buyer conversion likelihood score and optimal incentive package determination under various mortgage rate inputs.
Challenge input
JSON with base_rate_pct, buyer_target_price, buyer_max_monthly_budget, buyer_savings
AI SDK
Required framework for streaming web agents.
Qwen 3
Required LLM model engine.
Evaluated output
JSON with conversion_probability, recommended_incentive_type, incentive_cost, calculated_monthly_payment
- Verifies calculated payment is strictly within buyer_max_monthly_budget
- Accuracy of buyer conversion prediction based on budget fit • target: 0.92 • range: 0-1
- Benchmark: SalesConversionBench
- Conversion Predictor AUC target: 0.92
- 1 public reference case
- JavaScript execution harness
- javascript 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
- Vercel AI SDK · Optional
- Zed · Optional
- Qwen 3 · Required
- Vercel AI SDK · Optional
Evaluation contract
- Budget Constraint Verification · Weight 1
- Conversion Predictor AUC · Weight 1
Metrics
- Conversion Predictor AUC · Target 0.92
Benchmarks
- SalesConversionBench
- general
Reference cases
- 1 public reference case
Workflow tasks
- Task 1
Input: JSON with base_rate_pct, buyer_target_price, buyer_max_monthly_budget, buyer_savings
Output: JSON with conversion_probability, recommended_incentive_type, incentive_cost, calculated_monthly_payment
Execution harness: JavaScript
Evaluation tools
- AI SDK · Agent Framework
- Qwen 3 · Inference Model
Recipe freeze marker
No recipe freeze time is recorded.