Key Concepts
Projects
Section titled “Projects”Projects are containers for related notebooks. You might create one project per research question, experiment series, or collaboration. Each project has its own AI memory, so the assistant’s saved knowledge stays scoped to the relevant context. When you chat at the project level, the assistant can search across all notebooks in that project.
Notebooks
Section titled “Notebooks”Notebooks are ordered lists of blocks — like pages in a lab notebook. Each notebook lives inside a project and captures a single thread of work: an analysis, a protocol, a literature review. You can have as many notebooks as you need within a project, and the AI assistant can read and reference any of them.
Blocks
Section titled “Blocks”Blocks are the building units of a notebook. There are four types:
- Markdown — Rich text with support for headings, lists, links, and LaTeX math (rendered with KaTeX). Use these for notes, hypotheses, documentation, and narrative.
- Code — Python code that runs in your browser. Variables persist across code blocks within the same notebook, so you can build up an analysis step by step.
- Data Grid — Tabular data displayed in an interactive grid. Useful for viewing dataframes, CSV data, or structured results.
- Image — Figures, plots, and diagrams. Images generated by code blocks (like matplotlib plots) appear automatically.
AI Memory
Section titled “AI Memory”The AI assistant remembers key findings, decisions, protocols, and patterns that you approve for saving. Memory gives the assistant instant recall of important facts without needing to re-read entire notebooks. It is a shortcut, not a gate — the assistant can always search your notebooks directly. You control what gets saved, and you can review, delete, or archive memories at any time.