This article provides best practices for ingesting IoT device data using AWS IoT Core and Amazon Kinesis, helping organizations choose the right service based on their specific requirements.

• AWS IoT Core: Manages billions of devices, supports MQTT, enables command-and-control, fine-grained access control
• Amazon Kinesis: Processes high-throughput streaming data cost-effectively, scales with serverless architecture
• Consider device environment: fleet size, location, software stack, and operational control needs
• Evaluate activity: data volume/rate, communication flow, protocol requirements, and data purpose
• Define security posture: X.509 certificates for uncontrolled environments, AWS credentials for controlled ones
• MQTT supports payloads up to 128KB; Kinesis Data Streams handles up to 1MB payloads
• Combined approach: Use IoT Core for device management plus Kinesis for high-frequency data ingestion
• Decision matrix provided: IoT Core best for command-control and bidirectional communication; Kinesis best for high-throughput

Choose services based on payload size, throughput requirements, device constraints, and whether bidirectional communication or fine-grained access control is needed.