Software and Platforms I Use
A curated collection of tools and utilities I use or have evaluated.
| Tool | Category | Cost | Commentary |
|---|---|---|---|
| Ollama | Local LLMs | Free | Essential tool for running AI models locally. Great integration with other tools through its API. Active development and strong community support. |
| Open WebUI | AI Interface | Free | Clean interface for local AI models. Works seamlessly with Ollama. Customizable prompts and easy model switching. |
| Portainer | Docker Management | Free CE | Makes Docker container management visual and intuitive. Essential for managing multiple AI containers and services. |
| Neo4j | Graph Database | Free CE | Powerful graph database for AI knowledge storage. Great for implementing RAG systems and knowledge graphs. |
| ChromaDB | Vector Database | Free | Efficient vector storage for AI embeddings. Simple API and good performance for local development. |
| LangChain | AI Framework | Free | Essential framework for building AI applications. Great documentation and active community support. |
| Cloudflare Workers | Cloud / Edge Platform | Free tier | My primary deploy target — serverless edge compute with Workers AI, D1, R2, KV, Durable Objects, and Vectorize under one platform. |
| Anthropic Claude API | Model Provider | Usage-based | Primary LLM API for agents and apps. Strong tool use, long context, and prompt caching. |
| Claude Code | AI Coding Agent | Usage-based | Terminal-native coding agent I run daily for building, refactoring, and reviewing across repos. |
| OpenAI API | Model Provider | Usage-based | GPT models, embeddings, Whisper, and the Agents/Apps SDKs. |
| Groq | Fast Inference | Free tier | Ultra-low-latency inference for open models. Great for real-time agent loops; I also run the Groq desktop client. |
| Model Context Protocol (MCP) | Agent Protocol | Open standard | Open standard for connecting agents to tools and data. The backbone of most of my server and client projects. |
| LangGraph | Agent Framework | Free (OSS) | Graph-based orchestration for stateful, multi-step agents. My go-to for building deep agents and swarms. |
| LangSmith | Observability / Evals | Free tier | Tracing, evals, and prompt management for LLM apps. Indispensable for debugging agent runs. |
| Cursor | AI Code Editor | Freemium | AI-first editor built on VS Code with fast inline edits and codebase-aware chat. |
| VS Code | Code Editor | Free | My primary general-purpose editor and extension host. |
| Docker Desktop | Containers | Free / Paid | Local container runtime for building and testing services before they ship to the edge or a server. |
| Hono | Web Framework | Free (OSS) | Tiny, fast web framework that runs great on Workers. My default for edge APIs. |
| Playwright | Browser Automation | Free (OSS) | Reliable end-to-end browser automation and testing; also powers browser-use agent tooling. |
| Microsoft Azure | Cloud Platform | Free tier / Paid | Microsoft cloud, used via the Azure CLI and azd for hosting, Functions, and identity (Entra External ID). |
| Azure AI Foundry | AI Platform | Usage-based | Microsoft platform for building, evaluating, and deploying AI agents and models; I also run Foundry Local for on-device models. |
| Cloudflare Vectorize | Vector Database | Free tier | Edge-native vector store that pairs with Workers AI for low-latency RAG. |
| Hugging Face | Model Hub | Free tier | Models, datasets, and Transformers for local inference and fine-tuning. |
| Tavily | AI Web Search | Free tier | Search API built for LLMs; my default web-search tool for research agents. |
| Daytona | Dev Sandboxes | Free (OSS) | Fast, disposable dev environments and sandboxes for running agent-generated code. |
| Terraform | Infrastructure as Code | Free (OSS) | Declarative infra provisioning, including the Cloudflare provider. |
| Google Gemini | Model Provider | Free tier | Gemini models for multimodal tasks and Veo video experiments. |
| Obsidian | Knowledge Base | Freemium | Local-first markdown knowledge base where I keep notes and research. |