Awesome FDE Resources 
A focused list of high-quality resources for Forward Deployed Engineers and production AI delivery.
We prefer credible books, guides, production practices, real implementations, and tools that help move AI into real business workflows. Quality, relevance, maintenance, and practical value are reviewed together.
Contents
- Getting started - 19
- Best practices - 15
- Cases and reference implementations - 8
- Tools - 17
- FDE member network
- Hosted Chinese resources - 4
Getting started
- xdash/FDE-the-Guidance-Book-of-Forward-Deployed-Engineer - Chinese FDE guidance book covering discovery, activation, renewal, expansion, and scale.
- pierpaolo28/Awesome-FDE-Roadmap - Roadmap for FDE skills across agents, enterprise data, and strategic consulting.
- mlabonne/llm-course - Structured LLM learning path with notebooks and practical roadmaps.
- chiphuyen/aie-book - Supporting resources for Chip Huyen's AI Engineering.
- Hugging Face/agents-course - Hands-on course for understanding and building AI agents.
- ombharatiya/ai-system-design-guide - Production AI system design and evaluation guide.
- PacktPublishing/LLM-Engineers-Handbook - Practical LLM, RAG, deployment, and LLMOps handbook.
- decodingai-magazine/llm-twin-course - End-to-end production LLM and RAG course.
- MLOps-Courses/mlops-coding-course - Build and maintain a production MLOps codebase.
- paiml/practical-mlops-book - Code and exercises from Practical MLOps.
- Meirtz/Awesome-Context-Engineering - Survey of context engineering research, patterns, and implementations.
- curiousily/AI-Bootcamp - Self-paced GenAI bootcamp covering RAG, agents, and fine-tuning.
- ashishps1/learn-ai-engineering - Free learning path for AI and LLM engineering.
- microsoft/mcp-for-beginners - Cross-language introduction to Model Context Protocol.
- anmolksachan/AI-ML-Free-Resources-for-Security-and-Prompt-Injection - Beginner roadmap for AI security and prompt injection testing.
- dair-ai/MLOPs-Primer - Curated foundation for learning MLOps.
- libaice/Awesome-FDE - Bilingual collection of FDE resources, companies, core skills, and interview preparation materials.
- bryanyzhu/agentic-ai-system-course - Course skeleton for designing and operating production agents.
- alexeygrigorev/ai-engineering-field-guide - Data-backed guide to AI engineering roles, including responsibilities and skills extracted from 146 FDE job postings.
Best practices
- EthicalML/awesome-production-machine-learning - Libraries and practices for deploying, monitoring, versioning, and scaling ML.
- NirDiamant/agents-towards-production - Code-first tutorials from agent prototype to enterprise deployment.
- ai-boost/awesome-harness-engineering - Patterns for agent memory, permissions, evals, MCP, and observability.
- benchflow-ai/awesome-evals - Focused collection of agent evaluation resources and benchmarks.
- Puliczek/awesome-mcp-security - MCP security tools, research, and defensive guidance.
- requie/AI-Red-Teaming-Guide - Practical adversarial testing and AI security evaluation.
- vllm-project/guidellm - Evaluate LLM serving performance under real inference workloads.
- visenger/awesome-mlops - Curated MLOps references and production tooling.
- lizhe2004/Awesome-LLM-RAG-Application - RAG application patterns and implementation resources.
- suboss87/FDEOps - Local-first engagement memory, delivery methodology, and agent skills covering discovery through handoff.
- NVIDIA-NeMo/Guardrails - Programmable guardrails for LLM applications.
- GoogleCloudPlatform/agent-starter-pack - Production agent templates with CI/CD, evaluation, and observability.
- humanlayer/12-factor-agents - Twelve engineering principles for reliable customer-facing agent systems.
- databricks/databricks-agent-skills - Official skills encoding Databricks data, governance, evaluation, and deployment workflows for coding agents.
- anthropics/cwc-long-running-agents - First-party harness primitives and evaluator loops for reliable long-running Claude agents.
Cases and reference implementations
- patchy631/ai-engineering-hub - Real-world tutorials for agents, RAG, and LLM systems.
- NirDiamant/GenAI_Agents - More than 50 agent techniques and implementations.
- ikatsov/tensor-house - Enterprise AI notebooks for marketing, pricing, supply chain, and manufacturing.
- muratcankoylan/AI-Investigator - Adaptable enterprise research and report-generation system.
- GoogleCloudPlatform/generative-ai - Google Cloud GenAI samples and enterprise notebooks.
- GURPREETKAURJETHRA/END-TO-END-GENERATIVE-AI-PROJECTS - End-to-end GenAI projects with deployment examples.
- pathwaycom/llm-app - Ready-to-run RAG, live-data, and enterprise-search templates.
- Azure-Samples/AI-Gateway - Labs for AI gateways, models, MCP servers, and agents.
Tools
- langfuse/langfuse - LLM observability, evaluation, prompt management, and datasets.
- Arize-ai/phoenix - Open-source AI observability and evaluation.
- evidentlyai/evidently - Evaluation and monitoring for ML and LLM systems.
- promptfoo/promptfoo - Prompt, agent, and RAG testing with red teaming.
- confident-ai/deepeval - Evaluation framework for LLM applications.
- vibrantlabsai/ragas - Evaluation framework for RAG and LLM applications.
- mlflow/mlflow - Lifecycle, evaluation, tracing, and monitoring for AI systems.
- modelcontextprotocol/servers - Reference MCP servers and integrations.
- archestra-ai/archestra - Enterprise MCP registry, gateway, guardrails, and orchestration.
- langgenius/dify - Platform for agent workflows, RAG, models, and tools.
- daytonaio/daytona - Secure infrastructure for running AI-generated code.
- langchain-ai/langgraph - Runtime and orchestration framework for resilient agents.
- google/agents-cli - CLI and skills for creating, evaluating, governing, and deploying agents on Google Cloud.
- opensandbox-group/OpenSandbox - Secure, extensible sandbox runtime with SDK, MCP, Docker, and Kubernetes support.
- microsoft/PyRIT - Generative AI risk identification and red-team framework from Microsoft.
- langwatch/scenario - Multi-turn agent testing with simulated users, edge cases, and framework-agnostic evaluation.
- snyk/agent-scan - Security scanner for prompt injection and vulnerabilities in agents, MCP servers, and skills.
FDE member network
The member directory introduces practitioners and teams that can provide FDE knowledge, delivery capability, projects, or reusable resources. Entries require public evidence and clear conflict-of-interest disclosure.
Hosted Chinese resources
- FDE: A New Organizational Species in the AI Era - FDE Deployment in China Whitepaper 2026 (Preview) - A 15-page decision preview from Cui Niu Hui Research on deployment gaps, the role of FDE, and when enterprises need deep deployment.
- How to Become an FDE Engineer (Key Notes) - Zhang Zifeng ARK's concise notes on FDE capabilities, production deployment, deliverables, portfolio building, and customer discovery.
- FDE Interview: AI Deployment, Productization, and the Talent Profile - A structured 16-page editorial reconstruction of an interview covering enterprise AI delivery, Cresta's field practice, FDE hiring, trust, product feedback loops, and career paths.
- 100 Questions About Forward Deployed Engineers - AI Leishu's 92-page question-led guide to the FDE opportunity, role boundaries, skills, customer work, delivery, commercialization, and long-term career development.
These files remain under their named authors' and publishers' rights. See each neighboring .source.yml record for provenance and permitted use.
Contributing
Read CONTRIBUTING.md. New submissions should explain their direct value to FDE or production AI delivery. Emerging projects with strong evidence are welcome.
Linked projects remain under their owners' licenses. Original content is licensed under CC BY 4.0; code under scripts/ is licensed under Apache-2.0. See licensing details.