AI agent frameworks and platforms that I use to build practical systems for research, analysis, content, administration and other repeatable business tasks.
Pydantic AI is a Python framework I use to build structured AI agents and applications powered by large language models. It provides a developer-focused approach to connecting models, tools and data while keeping outputs structured and reliable, making it useful for building practical AI systems that integrate with wider business processes.
pydantic.dev
LangChain is a framework I use to build AI applications and agents that connect language models with tools, data and external systems. Its broad ecosystem supports more complex AI workflows involving reasoning, retrieval and automation, making it useful for developing practical applications that need to interact with multiple sources and business processes.
Dify is a visual platform I use to build, test and deploy AI applications, workflows and agents powered by large language models. Its low-code approach makes it useful for combining models, knowledge, tools and business logic into practical AI systems without requiring every part of an application to be developed from scratch.
LlamaIndex is a framework I use to connect AI applications and agents with documents, data and other knowledge sources. It is particularly useful for building retrieval and knowledge-based systems that allow language models to work with relevant business information, making it easier to create practical AI applications grounded in specific data and content.
