Resources

• 1) Platforms (Google / Microsoft / Others)
• 2) LLM & Model Ecosystem (Model Hubs + Providers)
• 3) NVIDIA Resources (Inference, Optimization, Agent Lifecycle)
• 4) Agent Frameworks & Orchestration
• 5) RAG & Vector Databases
• 6) Deployment & Inference Serving
• 7) Evaluation, Testing, and Observability
• 8) Guardrails, Safety, and Governance
• 9) Learning & Reference Materials

• Google AI Studio: https://aistudio.google.com/
• Use for rapid prototyping and prompt/agent behavior exploration.

• Vertex AI Platform overview: https://cloud.google.com/vertex-ai
• Vertex AI documentation: https://docs.cloud.google.com/vertex-ai/docs
• Vertex AI Studio (console prototyping): https://cloud.google.com/generative-ai-studio
• Use for managed model endpoints, enterprise deployment patterns, and integrations.

• Copilot Studio overview: Copilot Studio Overview
• Copilot Studio documentation: https://learn.microsoft.com/en-us/microsoft-copilot-studio/
• Great for governance, connectors, enterprise deployment, and agent management.

• Hugging Face docs: https://huggingface.co/docs
• Model hub (browse models): https://huggingface.co/models
• LLM Course (free learning resource): Hugging Face LLM Course

• Gemini API docs: https://ai.google.dev/
• Migration notes (Gemini API ↔ Vertex AI): Migrate to Vertex AI

Use for understanding model capabilities, limits, and integration options.

• Open-source models are commonly discovered via Hugging Face and vendor hubs.
• When evaluating models, check: licensing, context length, tool calling support, latency/cost, and safety features.

In this course, the “best model” is usually task-dependent. Focus on requirements and evaluation.

• NIM documentation: https://docs.nvidia.com/nim/index.html
• NIM developer portal: https://developer.nvidia.com/nim

Use to understand production-grade model serving on NVIDIA-accelerated infrastructure.

• NVIDIA docs hub (NeMo + more): https://docs.nvidia.com/

Use for agent lifecycle patterns, training/fine-tuning workflows, and enterprise tooling.

• Flowise GitHub: https://github.com/FlowiseAI/Flowise
• Use in this course for transparent, visual agent workflows.

• LangChain docs: https://docs.langchain.com/
• Hugging Face provider integration example: LangChain × Hugging Face

• LlamaIndex docs: https://docs.llamaindex.ai/

• Microsoft AutoGen (agentic workflows): https://microsoft.github.io/autogen/
• CrewAI (multi-agent coordination): https://docs.crewai.com/

We may reference these patterns conceptually; FlowiseAI remains the main course scaffold.

• Pinecone: https://www.pinecone.io/
• Weaviate: https://weaviate.io/
• Milvus: https://milvus.io/
• Chroma: https://www.trychroma.com/

Choose based on scale, latency, filters/metadata, and operational complexity.

• vLLM: https://docs.vllm.ai/
• KServe (Kubernetes model serving): https://kserve.github.io/website/
• Ray Serve: https://docs.ray.io/en/latest/serve/

Use when you need scalable endpoints, batching, and production reliability.

• LangSmith (LangChain tracing/eval): https://www.langchain.com/langsmith
• Arize Phoenix (LLM observability/eval): https://phoenix.arize.com/
• TruLens (evaluation framework): https://www.trulens.org/

Use for debugging agent loops, measuring quality, and catching regressions.

• NVIDIA NeMo Guardrails: https://github.com/NVIDIA/NeMo-Guardrails
• Guardrails AI: https://www.guardrailsai.com/

Use for policy constraints, safer outputs, and tool-call validation.

• Hugging Face LLM Course: LLM Course
• Vertex AI docs: Vertex AI Documentation
• Copilot Studio docs: Copilot Studio Documentation
• NVIDIA NIM docs: NVIDIA NIM Documentation

If you are unsure where to begin, start with the tools you are using this week in class.