Skip to main content
← All Patterns

Pattern

Arc

Efficient connection between points. One-directional sync with minimal transformation. The shortest path that works.

Definition

An Arc is the minimal viable connection between two systems. Not bidirectional synchronization. Not complex transformation pipelines. Just data flowing efficiently from A to B.

The name comes from geometry: an arc is a portion of a curve—the most direct path between two points on a circle. In systems design, an Arc connects two services with the least complexity required to achieve the goal.

"The Arc pattern asks: what's the shortest path between these two points that actually works in production?"

Principles

One Direction

Data flows one way. Source to destination. No round-trips, no sync conflicts, no reconciliation logic. Simplicity through constraint.

Gmail → Notion (not Gmail ↔ Notion)

API → Database (not real-time bidirectional)

Event → Handler (fire and forget)

Minimal Transformation

Transform only what's necessary. Preserve source fidelity. Don't normalize data that doesn't need normalizing.

✓ Map fields directly when schemas align

✓ Preserve formatting (links, bold, structure)

✗ Don't build complex ETL when simple mapping works

Serverless Infrastructure

Arcs should be stateless. No servers to maintain. Run on-demand via scheduled triggers or webhooks. Pay only for what you use.

✓ Cloudflare Workers, AWS Lambda

✓ Cron triggers for polling patterns

✓ Webhook endpoints for push patterns

OAuth over API Keys

For multi-user systems, OAuth provides proper authorization. Each user authenticates with their own credentials. No shared secrets.

✓ User-scoped access tokens

✓ Automatic token refresh

✓ Revocable permissions

When to Use

Good Fit

  • • Data capture from one system to another
  • • Automated backups and archiving
  • • Event logging and analytics pipelines
  • • Notification forwarding
  • • Report generation from live data

Poor Fit

  • • Real-time collaboration features
  • • Bidirectional sync requirements
  • • Complex data reconciliation
  • • Systems requiring strong consistency
  • • Interactive user workflows

Reference Implementation

Arc for Gmail

Gmail → Notion sync for Half Dozen

~11 hours to production

99.8% cost savings vs manual

Multi-user OAuth integration syncing labeled Gmail threads to a Notion database. 5-minute polling cycle, zero production failures, automatic contact creation.

5 min
Sync cycle
0
Production failures
$6.30
AI development cost

Architecture

Gmail API (OAuth) → Cloudflare Worker → Notion API

KV Store for token persistence and sync state

Cron trigger every 5 minutes