Architecture

The 2026 Unofficial Composable Deezwatts Stack

This year is more on the edge… and less reliant on the Rivian API.

The technical architecture for “DEEZWATTS: A Rivian Data Adventure” depicts a sophisticated, hybrid edge-to-cloud data pipeline designed for real-time vehicle telemtry and long-term analysis. At the edge, the system utilizes a Raspberry Pi running Ubuntu as an onboard compute unit within the vehicle, referred to in the diagram as “The Whip.” This hardware manages containerized workloads, or pods, to handle localized data processing. Connectivity is established through two distinct channels: a standard integration with the Rivian API for official telemetry and a decentralized Meshtastic Geo Mesh/LoRa network, which likely provides off-grid positioning and communication capabilities in areas without cellular service.

The middleware and infrastructure layer transition these edge signals into a managed environment where Kubernetes provides the necessary network and compute orchestration. Development and creative collaboration are centered on GitHub, facilitating version control for the project’s codebase. For the data’s final destination, the architecture leverages Google Cloud Platform to create a durable persisted store. By routing processed information into BigQuery, the system enables high-performance analytical queries and historical data warehousing, ultimately feeding into a specialized data visualization suite represented by the project’s custom branding.