Which template to start from
A project in the workbench is one body of research: the sources you read, the data you collect, and everything you write from them. When you create a project you choose what it starts from, and that gives it its first folders.
Templates are starting points, not rules. Every folder can be renamed, moved or deleted, new folders can be added at any depth, and the template a project began with is recorded but never enforced.
The workbench is still being built. Parts marked Built exist in the local workbench, which is not yet released; Being built is under way now, and Planned is designed and scheduled but not yet written.
One project for one body of research
The choice that matters most, and the easiest one to get wrong early on.
Returning to the same sources? One project.
Everything you read is imported once and annotated once, and every piece of writing in the project can draw on it.
Splitting a project splits its notes
A file imported into two projects is stored only once, but it becomes two documents, and notes and annotations belong to the project they were made in. A project per paper leaves your reading notes scattered across all of them.
A new project for genuinely separate work
A collaboration with a different team, a commissioned review, a teaching module: work whose sources you will not keep coming back to from your main research.
A doctorate: several papers, one thesis
MPhil or PhD, full-time over three or four years or part-time over as many as seven.
Start from Doctorate (MPhil/PhD), and make it the only project for the whole doctorate. The transfer report, each paper published along the way and the thesis at the end are all written from the same literature and the same data, so they belong in the same project. Moving from MPhil to PhD registration does not change the project; rename it if you like. Built
Reading and positioning
Import sources into Literature Review, then read, annotate and tag them. Each source is imported once however many papers later cite it: importing the same file again records where it came from instead of creating a duplicate. Being built
Design, ethics and data
Methods in Methodology, applications and approvals in Ethics & Approvals, collected material in Data. If the research involves participants, mark the folder that holds their material as human-subject data: everything in it gets the strictest level of protection, and moving it somewhere else never lowers that protection. Built Linking material to participants' consent, and restricting which AI providers may process it, are Planned.
Milestones and the transfer report
Progress reviews, annual reports and the MPhil-to-PhD transfer report go in Milestones. The transfer report is usually the first long piece of writing, and it draws on literature you have already annotated in the same project. Built
Papers along the way
Each paper is written inside the project, from its sources. A paper is made of sections in order, with a contents list made for you, and each section can be marked as ready; the mark is removed by itself if the text changes. When you seal a version, the workbench keeps a copy of every section exactly as it was, so years later you can still show which text your supervisor read. A sealed version can be saved as a Word document to send. What comes back — submissions, reviewers' comments, your responses — goes in Submissions & Reviews. Built
Supervision
Meeting notes, feedback and agreed actions in Supervision. A supervisor can be given the whole project or a single paper version: as a package they open with no account, with their comments coming back anchored to the passage they commented on, or shared live through the cloud workbench. Planned
The thesis
The thesis is a paper like the others, written in Thesis. Chapters often grow from published papers: comparing a chapter with its paper shows what changed in the text, the references and the ideas. The workbench can also read each cited source again, to make sure a chapter never claims more than its sources support. Planned
The examination
With every paper and the thesis in one project, you can show which sources several of them rely on, where two of them read the same source differently, and — for any help an AI gave — which model worked on which source and which of its suggestions you accepted. Planned
Every starting point
When to choose each template, and the folders it creates.
Doctorate (MPhil/PhD)
A doctorate over several years, with papers written along the way and a thesis at the end, all drawn from the same sources and data.
- Literature Review
- Methodology
- Ethics & Approvals
- Data
- Supervision
- Milestones
- Submissions & Reviews
- Thesis
- Notes & Ideas
One project for the whole doctorate, from MPhil registration to the examination.
Master's thesis or dissertation
One thesis or dissertation, usually written over a year or two, without papers expected along the way.
If a master's by research becomes a doctorate, carry on in the same project and add the folders the doctorate needs.
Taught module
A module on a taught course: its readings and lectures, and the essays, proposals and coursework it is assessed by.
One project per module. Each essay, proposal or piece of coursework is a paper in it, written in the workbench or in Word and checked in.
Literature review
A review that stands on its own: a review article, a scoping exercise, or a question to understand before deciding what to research.
Systematic review
A review that follows a protocol, with an explicit search, recorded screening decisions and data extracted from each included study.
Empirical study
Collecting data and analysing it to answer a question: experiments, surveys, measurements, observational data.
Qualitative study
Interviews, observation or documents, analysed for themes.
Interviews & Human Data starts marked as holding human-subject data, so everything filed in it gets the strictest level of protection.
AI-supervision study
Research on AI-assisted work itself: experiments across models and conditions, recorded agent runs, and the traces they leave.
- Literature Review
- Methodology
- Experiments
- Trace Models
- Coding Standards
- Agent Runs
- Interviews & Human Data
- Submissions & Reviews
- Notes & Ideas
- Diagrams & Architecture
The folders of the doctorate the workbench was first built for.
Blank project
None of these fits. Start empty and build the structure as the work takes shape.
No folders.
Select a folder to see what it is for, or read what each folder is for from the beginning.
Deciding quickly
If you are writing…
- a doctorate, with or without papers along the way
- Doctorate (MPhil/PhD)
- a master's thesis or dissertation
- Master's thesis or dissertation
- a module on a taught course: its essays, proposals and coursework
- Taught module
- a review article, or a review that stands on its own
- Literature review
- a review that follows a registered protocol
- Systematic review
- about data you collect and analyse
- Empirical study
- from interviews, observation or participants' own words
- Qualitative study
- about how people work with AI
- AI-supervision study
- something none of these describes
- Blank project
Part of a bigger piece of research? A review or a study that belongs to your doctorate goes inside the doctorate project rather than in a project of its own. Create the folders it needs there, and its notes stay with the rest of your research.
Filling it as you read: saving pages from your browser into the folder they belong in.
On more than one computer: syncing your computer with the cloud, and the workbench in the cloud itself.