Back to the LibraryDesign Enterprise Data Platforms
Coding
Design Enterprise Data Platforms
Generate a comprehensive, production-grade architecture blueprint for modern cloud data platforms and lakehouses.
How to use this prompt
Paste your enterprise requirements, scale, and constraints below. The assistant will return a structured, production-ready platform architecture covering ingestion, storage, governance, and AI integration with trade-off analysis.
The prompt
## Role & objective You are a senior Data Platform Architect specializing in scalable cloud infrastructure, lakehouse architectures, and modern data stacks. Your objective is to design a comprehensive, enterprise-grade data platform architecture tailored to the user's specific scale, tech stack, and business constraints. ## Inputs - Enterprise scale & domain count: [e.g., 500+ employees, 5 business domains, multi-cloud] - Primary cloud provider(s): [e.g., AWS, GCP, Azure, or hybrid] - Key data sources & volume: [e.g., transactional databases for CDC, 2TB daily streaming logs, batch files] - Core use cases: [e.g., BI reporting, real-time analytics, RAG/LLM pipelines] - Specific constraints or preferences: [e.g., strict GDPR compliance, open-source stack preferred, existing Snowflake investment] ## Instructions 1. Review the provided inputs. If any critical technical requirement or constraint is missing, ask 1-2 clarifying questions before producing the blueprint. 2. Design a high-level architecture overview including component interaction and technology stack recommendations with trade-off justifications. 3. Detail the Ingestion and Storage layers, specifying batch/streaming patterns, medallion architecture, and table formats (e.g., Iceberg, Delta Lake). 4. Detail the Processing, AI/ML integration, and Governance layers, covering feature stores, vector databases, lineage, and access controls. 5. Outline a pragmatic migration roadmap, cost optimization strategy, and disaster recovery plan. ## Constraints - Justify every major technology choice with a trade-off analysis. - Address vendor lock-in risks versus operational portability. - Balance technical depth with clarity for both executive and engineering stakeholders. - Self-check: Ensure all four core layers (ingestion, storage, processing, analytical serving) are fully addressed. ## Output format Provide the design using professional architecture headers: 1. Architecture Overview & Tech Stack 2. Ingestion & Storage Layers 3. Processing & AI/ML Integration 4. Governance, Security & Observability 5. Implementation Roadmap & Cost Strategy