Talk
Stop Prototyping, Start Shipping
Libraries for Python, Node.js, and Golang, organized by the topics covered in the talk. Switch language below.
Choosing the Right Model
-
OpenAIhttps://github.com/openai/openai-pythonOfficial Python client for OpenAI API (GPT-4, GPT-3.5, etc.)
-
Google Generative AIhttps://github.com/google/generative-ai-pythonPython SDK for Google's Gemini models (Gemini 2.0 Flash, etc.)
-
Anthropic SDKhttps://github.com/anthropics/anthropic-sdk-pythonOfficial Python client for Claude models
-
LangChainhttps://github.com/langchain-ai/langchainFramework for building LLM applications with multiple model support
-
LlamaIndexhttps://github.com/run-llama/llama_indexData framework for LLM applications with model abstraction
Prompt Engineering & Context Setting
-
LangChain Prompt Templateshttps://python.langchain.com/docs/modules/model_io/prompts/Prompt templates and management for consistent context setting
-
Promptifyhttps://github.com/promptslab/PromptifyPython library for prompt engineering and optimization
-
Guidancehttps://github.com/guidance-ai/guidanceA guidance language for controlling large language models
-
PromptToolshttps://github.com/patrickloeber/prompttoolsOpen-source tools for testing and experimenting with prompts
Prompt Security & Message Types Separation
-
OpenAI Python (System Messages)https://platform.openai.com/docs/guides/prompt-engineeringOfficial documentation on using system/user/assistant message types
-
Guardrails AIhttps://github.com/guardrails-ai/guardrailsPython framework for building safe, reliable LLM applications
-
LLM Guardhttps://github.com/protectai/llm-guardSecurity toolkit for LLM applications with input/output validation
-
Prompt Injection Detectionhttps://github.com/vercel/ai/tree/main/packages/aiBest practices for prompt injection prevention (Python examples)
Content Generation
-
LangChain Chainshttps://python.langchain.com/docs/modules/chains/Chain different LLM calls for complex content generation workflows
-
Haystackhttps://github.com/deepset-ai/haystackEnd-to-end framework for building production-ready LLM applications
-
AutoGPThttps://github.com/Significant-Gravitas/AutoGPTAutonomous GPT-4 experimentation platform
Choosing the Right Model
-
OpenAI Node.js SDKhttps://github.com/openai/openai-nodeOfficial Node.js client for OpenAI API (GPT-4, GPT-3.5, etc.)
-
@google/generative-aihttps://github.com/google/generative-ai-nodejsNode.js SDK for Google's Gemini models (Gemini 2.0 Flash, etc.)
-
@anthropic-ai/sdkhttps://github.com/anthropics/anthropic-sdk-typescriptOfficial TypeScript/JavaScript SDK for Claude models
-
LangChain.jshttps://github.com/langchain-ai/langchainjsTypeScript/JavaScript framework for building LLM applications
-
Vercel AI SDKhttps://github.com/vercel/aiFull-featured SDK for building AI applications with React, Vue, and more
Prompt Engineering & Context Setting
-
LangChain.js Promptshttps://js.langchain.com/docs/modules/model_io/prompts/Prompt templates and prompt management for consistent context
-
Promptablehttps://github.com/cfortuner/promptableTypeScript library for building and managing prompts
-
AI SDK Prompt Helpershttps://sdk.vercel.ai/docs/guides/providers/openaiUtilities for prompt engineering and context management
-
Promptfoohttps://github.com/promptfoo/promptfooTesting and evaluation framework for LLM prompts
Prompt Security & Message Types Separation
-
OpenAI Node.js (System Messages)https://platform.openai.com/docs/guides/prompt-engineeringOfficial docs on system/user/assistant message types in Node.js
-
AI SDK Securityhttps://sdk.vercel.ai/docs/guides/securitySecurity best practices and prompt injection prevention
-
LLM Guard (Node.js)https://github.com/protectai/llm-guardSecurity toolkit with Node.js bindings for input/output validation
-
Heliconehttps://github.com/Helicone/heliconeObservability and security monitoring for LLM applications
Content Generation
-
LangChain.js Chainshttps://js.langchain.com/docs/modules/chains/Chain LLM calls for complex content generation workflows
-
AI SDK Streamshttps://sdk.vercel.ai/docs/reference/ai-sdk-core/stream-textStreaming and content generation utilities
-
Inngesthttps://github.com/inngest/inngest-jsWorkflow engine for building reliable LLM-powered applications
Choosing the Right Model
-
go-openaihttps://github.com/sashabaranov/go-openaiUnofficial but popular Go client for OpenAI API (GPT-4, GPT-3.5, etc.)
-
google/generative-ai-gohttps://github.com/google/generative-ai-goOfficial Go client for Google's Gemini models (Gemini 2.0 Flash, etc.)
-
anthropic-gohttps://github.com/anthropics/anthropic-sdk-goOfficial Go SDK for Claude models
-
go-llmhttps://github.com/tmc/langchaingoLangChain implementation in Go for building LLM applications
-
ollamahttps://github.com/ollama/ollama-goGo client for running local LLMs with Ollama
Prompt Engineering & Context Setting
-
langchaingo Promptshttps://github.com/tmc/langchaingo/tree/main/promptsPrompt templates and management utilities in Go
-
go-prompthttps://github.com/c-bata/go-promptBuilding interactive prompts (CLI-focused but useful for context)
-
go-openai Chat Completionhttps://github.com/sashabaranov/go-openai#chat-completionMessage-based chat completion with system/user/assistant roles
Prompt Security & Message Types Separation
-
go-openai Message Typeshttps://pkg.go.dev/github.com/sashabaranov/go-openai#ChatMessageProper message type separation (System, User, Assistant) in Go
-
google/generative-ai-go Safetyhttps://pkg.go.dev/google.generative.ai/go/v2#SafetySettingSafety settings and content filtering for Gemini API
-
go-sanitizehttps://github.com/kennygrant/sanitizeInput sanitization library for preventing prompt injection
-
langchaingo Memoryhttps://github.com/tmc/langchaingo/tree/main/memoryMemory management for maintaining context and conversation state
Content Generation
-
langchaingo Chainshttps://github.com/tmc/langchaingo/tree/main/chainsChain different LLM operations for complex workflows
-
langchaingo Agentshttps://github.com/tmc/langchaingo/tree/main/agentsAgent framework for autonomous LLM-powered applications
-
go-openai Streaminghttps://github.com/sashabaranov/go-openai#streamingStreaming responses for real-time content generation