UnifyAI tool stack

Claude

Claude is Anthropic's AI assistant, strong at reading comprehension and working with long documents. You use Claude in three ways: in chat for daily knowledge and writing work, as Claude Code for development, and as Claude Cowork — an AI agent that autonomously executes tasks on your computer.

Free AI scan Book intake Back to tool stack

The key features

What you can do with Claude

Four things you benefit from immediately — in plain language.

Claude in chat & Projects

Daily knowledge and writing work: asking questions, summarising and drafting texts — with your own documents as memory.

Understands long documents

Read, summarise and analyse contracts, reports and policy documents in one go, with nuanced answers.

Claude Code — for development

Runs in the terminal and gives Claude access to your codebase: build, migrate and review faster, human stays in control.

Claude Cowork — agent for knowledge work

A desktop AI agent that autonomously executes multi-step tasks on your files and apps — delivering a finished result.

What you use it for

In chat (and in Projects, where you store your own documents) Claude reads, summarises and analyses long texts in one go: contracts, reports and policy documents. You ask questions about your own knowledge and get reliable, nuanced answers — and Claude helps with writing and rewriting, from email to proposal.

Claude Code is the variant for technical teams. It runs in the terminal and gives Claude access to read your codebase, write files and execute commands. This lets you build, migrate and review software much faster — with a human checking every change.

Claude Cowork brings that same capability to knowledge work, without the terminal. It is a desktop AI agent that executes multi-step tasks from start to finish: summarising research, drafting documents, organising files. You grant Claude access to specific folders and tools (such as Google Drive and Gmail), and you get a finished result rather than just an explanation.

How UnifyAI uses this

We use Claude as the "thinking layer" in agents and workflows: for analysis, summarisation and Q&A over your own documents. For sensitive data we decide together what is processed where, with EU-aware choices and human oversight.

For technical projects we build with Claude Code — prototypes, migrations and reviews — always with a human checking the outcome before it goes live. And we help SME teams get started with Claude in chat and Cowork, so you can use it yourself in your daily work.

We configure every tool with GDPR awareness: we determine together which data is processed where, with EU hosting where possible and human oversight on the result.