Skip to content

References: Advanced Data, Emerging AI, and Autonomous Architectures

Curated sources for deeper study of data mesh, data lakehouse, lambda and kappa architectures, AI security, federated learning, edge AI, online learning, A/B testing architecture, and autonomous system design.

Books

  • Dehghani, Zhamak. (2022). Data Mesh: Delivering Data-Driven Value at Scale. O'Reilly Media. The definitive reference for data mesh architecture, covering the four principles of domain ownership, data-as-product, self-serve infrastructure, and federated governance that this chapter analyzes for modifiability and interoperability quality attribute tradeoffs.

  • Kleppmann, Martin. (2017). Designing Data-Intensive Applications. O'Reilly Media. Provides the foundational analysis of lambda and kappa architectures, event sourcing, and stream processing tradeoffs that this chapter examines through the ATAM lens of latency-consistency and operational complexity tradeoffs.

  • McMahan, H. Brendan, and Daniel Ramage. (2017). "Federated Learning: Collaborative Machine Learning without Centralized Training Data." Google AI Blog. While a blog post, this Google research communication introduced federated learning to the ML architecture community and provides the conceptual foundation for the federated learning quality attribute analysis in this chapter.

Articles and Papers

  • Armbrust, Michael, et al. (2021). "Lakehouse: A New Generation of Open Platforms that Unify Data Warehousing and Advanced Analytics." Proceedings of CIDR 2021. https://www.cidrdb.org/cidr2021/papers/cidr2021_paper17.pdf The foundational paper introducing the data lakehouse concept, open table format architecture (Delta Lake, Iceberg, Hudi), and the quality attribute improvements over traditional data warehouse and data lake architectures.

  • Biggio, Battista, and Fabio Roli. (2018). "Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning." Pattern Recognition, 84. Peer-reviewed survey of adversarial machine learning attacks (adversarial examples, model poisoning, model extraction) and defenses, providing the academic foundation for this chapter's AI security architecture analysis.

Online Resources

  • "Data Mesh Architecture." Thoughtworks. https://www.thoughtworks.com/en-us/radar/techniques/data-mesh Thoughtworks Technology Radar's coverage of data mesh as an organizational and technical approach, including real-world adoption experiences and the organizational readiness factors that this chapter identifies as ATAM risks.

  • "Apache Kafka Documentation: Streams." Apache Software Foundation. https://kafka.apache.org/documentation/streams/ Official documentation for Kafka Streams, the core technology behind kappa architecture's single-pipeline approach, providing the technical basis for the lambda vs. kappa operational complexity tradeoff analyzed in this chapter.

  • "Federated Learning." Google AI. https://ai.googleblog.com/2017/04/federated-learning-collaborative.html Google AI's original federated learning blog post and subsequent research links, covering the privacy-accuracy-communication tradeoffs that this chapter examines as ATAM quality attribute tensions in privacy-constrained training scenarios.

  • "Edge Impulse Documentation." Edge Impulse. https://docs.edgeimpulse.com Documentation for edge AI model optimization techniques (quantization, pruning, distillation) and deployment to constrained devices, supporting this chapter's analysis of the edge-cloud compute tiering pattern and quality attribute tradeoffs.

Videos

  • "Autonomous Agents and the Future of Software Architecture." Simon Wardley. GOTO Conference. YouTube. Analysis of emerging autonomous system architecture patterns, safety tradeoffs, and the architectural discipline required for systems that take consequential real-world actions — the ATAM frontier examined in this chapter's autonomous system architecture section.