Lyt når som helst, hvor som helst

Dyk ned i over 1 million e- og lydbøger samt podcasts.

Over 1 million titler
Eksklusive titler + Mofibo Originals
Download og nyd titler offline
Opsig når som helst

Prøv nu

Cover for Designing Data-Intensive AI Applications: An Architect's Blueprint for AI Systems Under Real-World Scale, Latency, and Compliance Pressure

Designing Data-Intensive AI Applications: An Architect's Blueprint for AI Systems Under Real-World Scale, Latency, and Compliance Pressure

Forfatter
- Jordan O'Neal
Forlag
- PublishDrive

Serier: 5 af 2
Sprog: Engelsk
Format
Kategori: Fakta

Most enterprise AI projects fail not because of the model, but because the architecture underneath cannot handle scale, latency, and compliance. This blueprint gives architects the patterns to design AI systems that hold up in production and pass the audit.

Every enterprise AI program eventually hits the same wall. The demo worked in the conference room, with one user, a curated dataset, and no latency or compliance requirements. Then the team is asked to serve a thousand tenants across regulated industries, and the real problem appears. It was never the model. It was the architecture.

Designing Data-Intensive AI Applications is the architect's blueprint for crossing that gap. Its central claim is stated plainly: AI applications are data-intensive distributed systems, and the model is one component among several. The spine of the system is data, meaning how it is stored, replicated, partitioned, processed, retrieved, and kept consistent as the system scales and ages.

Eleven chapters work through the four pillars of Data, Model, Retrieval, and Integration: storage from lakes to vector stores, replication and multi-region consistency, partitioning and multi-tenancy, batch, stream, and online processing, RAG and agentic systems, inference economics, enterprise integration, reliability and failure modes, and governance as a structural property. A final chapter assembles five reference architectures and case studies.

Each chapter opens with a composite scenario from a regulated enterprise and closes with a Technical Checklist, Team Discussion questions, and Key Takeaways a team can carry straight into a design review. Tool-neutral and regulator-aware, this is Book 5 of the Serious AI Manager Series and the working text for architects responsible for enterprise AI that has to perform under real-world scale, latency, and compliance pressure.

Udgivelsesdato

E-bog: 17. maj 2026