Skip to main content
Back to challenges
Challenge

Build an Oman Power Plant Milestone Agent using LangChain & Vercel AI SDK

Managing complex engineering milestones for $683m combined-cycle power plant contracts requires multi-agent document checking. Build a LangChain and Vercel AI SDK pipeline to process EPC contracts, detect schedule slippage, and deliver structured dashboard updates.

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.

Leverage LangChain stateful graphs combined with Vercel AI SDK streaming endpoints to monitor power plant construction schedules and alert project directors.

Delivery guide

How work is evaluated

Evaluation

Validates the agent's extraction accuracy for milestone dates and delay responsibility assignment.

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

variance_calc_check

Ensure date variance calculation is mathematically exact

Binary check

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

Dimension 2

milestone_accuracy

Accuracy of extracted milestones • target: 0.98 • range: 0.9-1

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

  • Construct stateful LangGraph agents for multi-tier milestone analysis

  • Stream live milestone risk evaluations via Vercel AI SDK UI primitives

  • Extract EPC contract terms and compare against current construction progress

  • Generate automated variance analysis report summaries for stakeholders

Resources and assets

Reference links and supporting material

Dataset notes

Construction milestone reports, procurement receipts, and delay logs for Oman power generation projects.

Challenge setup

Validates the agent's extraction accuracy for milestone dates and delay responsibility assignment.

Challenge input

JSON containing contract timeline and actual progress

LangChain

State graph management for multi-stage contract processing

Vercel AI SDK

Streams dynamic project updates to management web UI

Evaluated output

JSON variance summary

Evaluation checks
  • Ensure date variance calculation is mathematically exact
  • Accuracy of extracted milestones • target: 0.98 • range: 0.9-1
Configured evaluation evidence
  • Benchmark: EPCScheduleBench
  • Milestone Accuracy target: 0.98
  • 1 public reference case
Configured runtime
  • 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

Action Space
  • Vercel AI SDK · Required
  • LangChain · Optional
  • Martian · Optional
Orchestration
  • Vercel AI SDK · Required
  • LangChain · Optional

Evaluation contract

  • variance_calc_check · Weight 1
  • milestone_accuracy · Weight 1

Metrics

  • milestone_accuracy · Target 0.98

Benchmarks

  • EPCScheduleBench
  • general

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON containing contract timeline and actual progress

    Output: JSON variance summary

    Execution harness: Python

Evaluation tools

  • LangChain · Agent Orchestration
  • Vercel AI SDK · UI Streaming

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build an Oman Power Plant Milestone Agent using LangChain & Vercel AI SDK