Build GraphQL API on AWS Amplify using AppSync, Cognito, DynamoDB, Storage and Push Notifications

Build an API in AppSync that uses approximately 10 types. Most of the relationships between the types are one-to-many, there are a couple of many-to-many relationships. The API will use DynamoDb for the backend, and Cognito for user authentication. Recommend best practice for integrating Appsync API with Cognito to query user custom attributes. For example, we need to decide whether custom user attributes will be stored in Cognito, or the an AppSync type. The Appsync API will also use Amplify Storage to store images and video, which will be part of the API schema. One of the types will include location data in the form of GPS coordinates. We'll need to design AppSync integration with ElasticSearch to support geo searches. Design mutations for creating data Design queries for retrieving data Design subscriptions for real-time publishing of changes Support client use of AWS DataStore for offline sync of data This is a draft of the current schema that we will use as a starting point: schema { query:Query } type Query { getSensors: [Sensor] } type Sensor @model { id: ID! image_url: String status: String battery: Int productName: String serialNumber: String } type Agency @model { id: ID! name: String! abbreviation: String! patchImage: Media # TODO: refactor stations: [Station] @connection(name: "AgencyStations") } type Station @model { id: ID! name: String! abbreviation: String! patchImage: Media # TODO: refactor address1: String! address2: String city: String state: String postalCode: String agency: Agency @connection(name: "AgencyStations") units: [Unit] @connection(name: "StationUnits") } type Unit @model { id: ID! name: String! # TO DO: add phone number for SMS stations: Station @connection(name: "StationUnits") officers: [Officer] @connection(name: "UnitOfficers") } type Officer @model { id: ID! firstName: String! middleInitial: String! lastName: String! badgeNumber: String idImage: Media # TO DO: refactor deviceToken: String custodies: [Custody] @connection(name: "OfficerCustodies") unit: Unit @connection(name: UnitOfficers) } type Custody @model { id: ID! sensor: Sensor @connection(fields: ["sensorID"]) sensorID: ID! ageUpper: Int ageLower: Int ageActual: Int gender: String currentOfficerID: String currentStatus: currentStatus officer: officer @connection(name: OfficerCustodies) # Current officer of record for custody media: [Media] # TO DO: refactor custodyState: [CustodyState] # TO DO: refactor custodyLocations: [Location] # TO DO: refactor officerOfRecord: [CustodyOfficerOfRecord] # TO DO: refactor } type CustodyOfficerOfRecord @model (queries: null) @key(name: "byOfficer", fields: ["officerID", "officerID"]) @key(name: "byCustody", fields: ["custodyID", "custodyID"]) { id: ID! officerID: ID! custodyID: ID! createdTime: Int! officer: Officer! @connection(fields: ["officerID"]) custody: Custody! @connection(fields: ["custodyID"]) } enum currentStatus { Active Inactive } type CustodyState { createdTime: Int! cp_priority: Int cp_state: String cp_type: String cp_message: String custodyStatus: CustodyStatus } type CustodyStatus { createdTime: Int! type: String timerType: String title: String eventUtcTimestamp: Int intervalTimeInSeconds: Int } type Location { createdTime: Int! longitude: Float! latitude: Float! } type Media @model { id: ID! createdTime: AWSTimestamp! type: MediaType! name: String! description: String! s3Bucket: String! s3Region: String! s3Key: String! } enum MediaType { Image Video }