Skip to main content
Back to challenges
Challenge

Newborn Genomic Screening Risk Classifier using AI SDK and FHIR Integration

Early detection of pediatric cancer risks requires stream-processing newborn genomic profiles into standardized FHIR diagnostic resources. Build an agentic pipeline using Vercel AI SDK to map target variants to pediatric cancer risk scores and generate valid FHIR DiagnosticReport payloads with zero schema validation errors.

Data ScienceHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Implement a real-time newborn genomic screening classifier using Vercel AI SDK and HL7 FHIR standard formatting.

Delivery guide

How work is evaluated

Evaluation

Evaluates FHIR resource schema compliance and genomic risk classification accuracy.

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

fhir_schema_pass

Verifies FHIR JSON conforms to DiagnosticReport schema specification.

Binary check

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

Dimension 2

schema_error_rate

Percentage of generated FHIR payloads with schema errors. • target: 0 • range: 0-1

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

  • Build streaming AI agents in TypeScript using Vercel AI SDK.

  • Parse newborn genomic screening variant panels for pediatric cancer risk loci.

  • Generate standardized HL7 FHIR DiagnosticReport JSON resources.

  • Validate FHIR output payloads against official HL7 schemas.

Resources and assets

Reference links and supporting material

Dataset notes

Synthetic newborn germline sequencing panel dataset annotated with pediatric cancer risk classifications.

Challenge setup

Evaluates FHIR resource schema compliance and genomic risk classification accuracy.

Challenge input

JSON with newborn ID and list of detected germline variants.

AI SDK

Vercel's framework for building type-safe LLM interfaces and structured streaming.

Evaluated output

JSON containing HL7 FHIR DiagnosticReport resource.

Evaluation checks
  • Verifies FHIR JSON conforms to DiagnosticReport schema specification.
  • Percentage of generated FHIR payloads with schema errors. • target: 0 • range: 0-1
Configured evaluation evidence
  • Benchmark: FHIR Genomics interoperability Benchmark
  • Schema Error Rate target: 0%
  • 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
  • Zed · Required
  • GPT-Engineer · Optional
  • Mistral 7B · Optional

Evaluation contract

  • fhir_schema_pass · Weight 1
  • schema_error_rate · Weight 1

Metrics

  • schema_error_rate · Target 0

Benchmarks

  • FHIR Genomics interoperability Benchmark
  • biotech

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with newborn ID and list of detected germline variants.

    Output: JSON containing HL7 FHIR DiagnosticReport resource.

    Execution harness: JavaScript

Evaluation tools

  • AI SDK · Agent Framework

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Newborn Genomic Screening Risk Classifier using AI SDK and FHIR Integration