Skip to main content
Loading navigation
Back to challenges

Hospitality Megaproject Construction Copilot with Vercel AI SDK & Aider

Real estate developers constructing multi-hotel portfolios require real-time project risk alerts and automated code generation for dashboard components. Build an intelligent web dashboard service with the Vercel AI SDK and refactor code via Aider to track room delivery schedules.

Business OperationsHosted by Vera
Sign in free to start

What you are building

Construct a full-stack Next.js and Vercel AI SDK service backed by Aider automated code refactoring.

How work is evaluated

Evaluation

Evaluates streaming endpoint latency and structural schema accuracy for hotel completion telemetry.

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

valid_zod_schema

Ensure delivered payload parses cleanly through required TypeScript interface.

Binary check

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

Dimension 2

time_to_first_token

Latency in milliseconds before the first response chunk is streamed. • target: 350 • range: 0-2000

Binary check

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

Learning goals

  • Implement multi-provider AI routing using Vercel AI SDK in Next.js App Router

  • Stream UI components and room milestone updates live to the client

  • Integrate Aider CLI workflow for automated prompt-driven code changes

  • Build dynamic schema validation for construction delivery milestones

Resources

Dataset notes

Data records detailing hotel development milestones, room counts, location breakdown, and target handover dates across Saudi cities.

Challenge setup

Evaluates streaming endpoint latency and structural schema accuracy for hotel completion telemetry.

Challenge input

JSON with portfolio_id and target_quarter

AI SDK

TypeScript framework for streaming object and text responses.

Aider

Automates code refactoring and maintenance.

Evaluated output

Stream of JSON objects containing total_rooms, delivered_rooms, and status

Evaluation checks
  • Ensure delivered payload parses cleanly through required TypeScript interface.
  • Latency in milliseconds before the first response chunk is streamed. • target: 350 • range: 0-2000
Configured evaluation evidence
  • Benchmark: FullStack-AI-Bench
  • Time To First Token target: 350
  • 1 public reference case
Configured runtime
  • JavaScript execution harness
  • javascript 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
  • Aider · Required
  • Vercel AI SDK · Optional
  • Synthflow · Optional
Observation
  • Aider · Required
Orchestration
  • Vercel AI SDK · Optional

Evaluation contract

  • valid_zod_schema · Weight 1
  • time_to_first_token · Weight 1

Metrics

  • time_to_first_token · Target 350

Benchmarks

  • FullStack-AI-Bench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with portfolio_id and target_quarter

    Output: Stream of JSON objects containing total_rooms, delivered_rooms, and status

    Execution harness: JavaScript

Evaluation tools

  • AI SDK · agent_framework
  • Aider · code_refactoring

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Hospitality Megaproject Construction Copilot with Vercel AI SDK & Aider