GenAI Architecture | FDE Interview Handbook

FDE Interview Handbook

Membership Discord Sign in / up

Fresh questions, updated knowledge · Sep 15, 2026

GenAI Architecture

1.1 From Traditional Software to GenAI Systems 1.2 The Spectrum of GenAI Applications 1.3 What Makes a System Agentic?1.4 Model + Harness 1.5 Vibe Coding vs. Agentic Engineering 1.6 The New Role of the Engineer

2.1 Foundation Models 2.2 Prompts and Instructions 2.3 Context 2.4 Knowledge and Retrieval Primitives 2.5 Structured Input and Output 2.6 Tools and Actions 2.7 Model APIs and Runtime 2.8 Model Context Protocol (MCP)2.9 State and Memory Primitives 2.10 Identity, Permissions, and Basic Guardrails

3.1 Pattern Selection: Start Simple 3.2 Direct LLM and Chatbot Pattern 3.3 RAG Architecture Pattern 3.4 Deterministic Workflow Pattern🔒3.5 Agentic Workflow Pattern🔒3.6 Skills and Dynamic Capability Loading 3.7 Multi-Agent Systems and A2A🔒3.8 Interoperability Interfaces 3.9 Human-in-the-Loop Patterns🔒3.10 Choosing the Right Pattern🔒

4.1 Why AI Reliability Is Different 4.2 Evaluation Fundamentals🔒4.3 Evaluating Different Architectures 4.4 AI Security and Threat Modeling🔒4.5 Runtime Guardrails and Recovery🔒4.6 Observability and Incident Evidence🔒4.7 Governance and Responsible AI 4.8 Reliability Readiness Checklist🔒

5.1 Spec-Driven Development 5.2 From Prototype to Production 5.3 Policy and Instruction Enforcement🔒5.4 Production Reference Architecture🔒5.5 Model Gateway and Provider Strategy 5.6 Deployment Platforms and Amazon Bedrock 5.7 Production Data, State, and Storage 5.8 AI-Native CI/CD and Versioning🔒5.9 Scaling, Performance, and Cost Engineering🔒5.10 Operations, Rollout, and Ownership🔒

loading…

FDE Interview Handbook

The complete prep guide for Forward Deployed Engineer interviews.

Learn

FDE Interview Guide
Study Plan
Commercial Mindset
Python Warmups
GenAI Architecture
Coding Practice
Blogs
Industry News
Tools

Contact

support@fdehandbook.com
Discord community

© 2026 FDE Interview Handbook