Skip to main content
Back to challenges
Challenge

AI SDK Tender Valuation Microservice on Seldon Core

Implement a Singapore URA land parcel valuation agent using Vercel AI SDK and Seldon Core to forecast land bid pricing. The pipeline aggregates urban planning parameters and outputs real-time bid yield curves within 200ms latency.

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.

Combines Vercel AI SDK TypeScript pipelines with Kubernetes-deployed Seldon Core inference graphs to evaluate urban land sales.

Delivery guide

How work is evaluated

Evaluation

Evaluates accuracy of land valuation predictions and inference latency.

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

Valuation Range Check

Verifies predicted bid is within reasonable historical Singapore land tender bounds

Binary check

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

Dimension 2

Mean Absolute Percentage Error (MAPE)

MAPE against historical URA winning tender prices • target: 4.5 • range: 0-15

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 TypeScript backend using Vercel AI SDK streaming and tool calling

  • Deploy model inference components onto Kubernetes using Seldon Core custom wrappers

  • Process Singapore URA Gross Plot Ratio (GPR) and site constraint metrics

  • Predict optimal bid price per square meter per plot ratio ($/psm/GPR)

Resources and assets

Reference links and supporting material

Dataset notes

Dataset of 40 past Singapore URA land sale releases including site specs, MRT distances, and historical winning bids.

Challenge setup

Evaluates accuracy of land valuation predictions and inference latency.

Challenge input

JSON containing site_location, site_area_sqm, gpr, estimated_units, and mrt_distance_m

AI SDK

Provides structured serverless pipeline orchestrations.

Seldon Core

Serves underlying containerized ML valuation microservices.

Evaluated output

JSON containing predicted_bid_psm_gpr, confidence_interval, and primary_value_drivers

Evaluation checks
  • Verifies predicted bid is within reasonable historical Singapore land tender bounds
  • MAPE against historical URA winning tender prices • target: 4.5 • range: 0-15
Configured evaluation evidence
  • Benchmark: UrbanLandValuation-APAC
  • Mean Absolute Percentage Error (MAPE) target: 4.5
  • 1 public reference case
Configured runtime
  • 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

Action Space
  • Seldon Core · Required
  • Vercel AI SDK · Optional
  • Capably · Optional
Orchestration
  • Vercel AI SDK · Optional

Evaluation contract

  • Valuation Range Check · Weight 1
  • Mean Absolute Percentage Error (MAPE) · Weight 1

Metrics

  • Mean Absolute Percentage Error (MAPE) · Target 4.5

Benchmarks

  • UrbanLandValuation-APAC
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing site_location, site_area_sqm, gpr, estimated_units, and mrt_distance_m

    Output: JSON containing predicted_bid_psm_gpr, confidence_interval, and primary_value_drivers

    Execution harness: JavaScript

Evaluation tools

  • AI SDK · TypeScript AI SDK
  • Seldon Core · Model Runtime

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about AI SDK Tender Valuation Microservice on Seldon Core