Query layer

GraphQL

Data LiveAPI Planned

A typed schema over the same graph, for teams that want field selection, code generation and a stable contract without writing traversals themselves.

Schema

Node kinds become types and edge kinds become fields, with the full schema served from the endpoint and readable by any standard client.

schema.graphql excerpt

type Candidacy {
  slug: ID!
  name: String!
  person: Person!
  contest: Contest!
  party: Party
  isIncumbent: Boolean
  filingStatus: FilingStatus
  ballotAccess: BallotAccess
  campaignWebsite: URL
  """Requires the `contact` entitlement."""
  campaignEmails: [EmailAddress!]!
  positions: [Position!]!
  """Every source this candidacy's fields were drawn from."""
  provenance: [Provenance!]!
  confidence: Float!
}

type Query {
  contests(state: StateCode!, level: Level, electionDate: Date): [Contest!]!
  candidacy(slug: ID!): Candidacy
  districtsAtPoint(lat: Float!, lng: Float!): [Region!]!
  ballotAtPoint(lat: Float!, lng: Float!, date: Date!): [Contest!]!
}

Example

POST /v1/graphql

query ArizonaCountyRaces {
  contests(state: "AZ", level: COUNTY, electionDate: "2026-11-03") {
    slug
    office { name level }
    region { name fips population }
    candidacies {
      person { name }
      party { name }
      isIncumbent
      confidence
    }
  }
}

Depth and complexity are budgeted. Queries are scored before they run and rejected above your plan's ceiling with an error naming the offending path, since heavy aggregate work belongs in Cypher or a bulk export instead.