@bradtech/datapoints

Quatrain-powered DataPoint domain model and repository for TimescaleDB / PostgreSQL timeseries ingestion and querying.

Part of the sovereign Brad Open Source Agronomic Stack (bradtech-oss).

🌟 Overview

@bradtech/datapoints provides the canonical domain entity and high-level repository for time-indexed telemetry, environmental metrics, and derived agronomic indicators in the Brad and Quatrain ecosystems.

• Domain Model (DataPoint): Extends Quatrain PersistedBaseObject, defining strongly-typed schemas, confidence scoring, and JSONB contextual metadata.
• Repository & Mini API (DataPointRepository): Extends BaseRepository<DataPointType>, using the fluent @quatrain/backend Query builder for filtering, time-range scanning, and point-in-time lookups.
• Backend Agnostic: Connects seamlessly with @quatrain/backend-postgres (TimescaleDB hypertables) or any standard Quatrain backend adapter.

📦 Installation

bun add @bradtech/datapoints @quatrain/core @quatrain/backend @quatrain/backend-postgres
# or
npm install @bradtech/datapoints @quatrain/core @quatrain/backend @quatrain/backend-postgres

🚀 Quick Start

import { Backend } from '@quatrain/backend'
import { PostgresAdapter } from '@quatrain/backend-postgres'
import { DataPointRepository, DataPoint } from '@bradtech/datapoints'

// 1. Initialize PostgreSQL / TimescaleDB Adapter
const postgres = new PostgresAdapter({
   config: {
      host: 'postgres.timeseries.svc.cluster.local',
      port: 5432,
      database: 'datapoints',
      user: 'brad_timeseries',
      password: process.env.DATAPOINTS_DB_PASSWORD,
   },
})
Backend.addBackend(postgres, '@default', true)

// 2. Instantiate Repository
const repo = new DataPointRepository(postgres)

// 3. Batch Ingest Telemetry
await repo.insertMany([
   {
      device: '8c1f645490100016',
      plot: 'plots/parcelle-nord',
      metric: 'okf:agronomy/soil/vwc_calibrated',
      value: 28.5,
      unit: '%',
      kind: 'computed',
      confidence: 0.98,
      timestamp: new Date().toISOString(),
   },
])

// 4. Query Timeseries Timeline
const timeline = await repo.getTimeline({
   plot: 'plots/parcelle-nord',
   metric: 'okf:agronomy/soil/vwc_calibrated',
   from: '2026-09-01T00:00:00Z',
   to: '2026-09-02T23:59:59Z',
   order: 'desc',
   limit: 50,
})

📖 Documentation

• 👉 HOWTO Guide (Common Use Cases)

📄 License

AGPL-3.0-or-later © 2026 Olivier Lépine & Brad Technology.