A clearer window on qualitative evidence.

Oriel runs on Windows and on Macs with Apple silicon (M1 or later), with a Linux version to follow. Download it here. Colleagues on Windows and Mac can share one project. Some features depend on your computer. Local AI currently needs an Apple silicon Mac or a Windows PC with a suitable graphics processor, and text recognition in scanned PDFs uses the engine built into Windows or macOS. See the platform notes.

Your working project is saved as a file on your computer. Coding, queries, reports, transcription, and optional AI processing run locally. If you choose collaboration or a cloud-synchronised folder, that folder or service also copies project material according to its access and storage arrangements.

Oriel’s optional AI runs on your computer after downloading a model. Your research material is not sent to a cloud AI service for processing.

No. You can code and analyse your material yourself. In the default suggestion mode, you review suggestions before accepting them. A separate, optional AI-coder mode saves a run directly under its own identity.

Yes. The trade-off for keeping AI processing on your computer is that it requires powerful hardware. Performance depends on your processor, available memory, graphics hardware, and chosen model. Less powerful computers may run models slowly or be unable to run larger ones. AI is optional, so you can use Oriel’s manual coding and analysis tools without it.

Yes. You can choose a separate AI coder for a run, inspect its coding, compare it with yours, and remove the run. The default mode gives you suggestions to accept or dismiss instead. Both approaches need your critical judgement.

Local processing helps you control where your material is processed. Before importing sensitive data, check your institution’s requirements, study approvals, and storage arrangements. A locally stored project may still be synchronised by services such as OneDrive.

Coding and analysis run locally. Download any transcription or AI models beforehand. Retrieving references from DOIs and capturing online sources require internet access.

Oriel imports NVivo .nvp and .nvpx projects, with a companion .qdpx export, a codebook, or both needed to restore names. Some information, including attribute values, memo text, and saved searches, does not transfer. It also supports REFI-QDA exchange and Word or PDF reports. Read any transfer warnings and check the result. Export back to NVivo isn’t fully supported.

Yes, through the storage you already use. Choose a folder in OneDrive, Teams or SharePoint, Dropbox, Google Drive, iCloud Drive, or a network drive, then share that folder with named colleagues as usual. Each person keeps a separate local copy outside the folder, works at full speed, including offline, and codes under their own name. There is no Oriel server or account. Everyone with access to the folder can read the whole project. See how it works or read the current limits.

It is designed to be. Your project is a single file on your computer, so opening sources, coding, and searching do not wait on a server. Text search uses an index, long PDFs draw only the pages near you, and large datasets scroll without loading every row at once. Coding is saved as you apply it, and ⌘K (Ctrl+K on Windows) opens any source, code, or command by name. Local AI and transcription speeds depend on your hardware.

No. You tell Oriel what your project is for, and your home page suggests five broad steps with the tools that suit them. Steps can stay empty and your questions can change. The counts describe your work rather than checking that you followed a particular method. See the route.

Yes. Choose Evidence synthesis when you create a project and Oriel will suggest Study review tools, or you can switch them on later. You can manage studies and their reports, share extraction and appraisal templates, reconcile two reviewers, link study contributions to findings, and record CERQual judgements. Oriel imports screening outcomes but doesn’t run searches, deduplicate records, or draw flow diagrams, and shared project files don’t technically blind reviewers. See the review workflow.

No. It supports general qualitative coding, matrices, rapid analysis using source summaries, and reporting, with additional tools for programme theory and CMOCs, behavioural and implementation science, and qualitative evidence synthesis.

Yes. Choose behavioural or implementation questions as part of an evaluation, or switch on Behavioural science, Implementation science, or both in any project. Oriel includes codes for the Theoretical Domains Framework, COM-B, the Behaviour Change Wheel, CFIR, ERIC, EPIS, Normalisation Process Theory, and RE-AIM, a TDF behavioural diagnosis, CFIR construct ratings with comparison of groups, logic models, a planning route for designing interventions, and records of strategies and adaptations. Counts help you see where your data has most to say, but every judgement and rating is yours. See the programme evaluation workflow.

Oriel is intended to be updated and improved over time. You are welcome to report bugs and request features. Justin develops Oriel as a passion project alongside his main work, so personalised support, response times, and fixes or requested features cannot be guaranteed.

Yes. AI was used to help develop the application. This means that the code upon which the programme was built was written with the assistance of AI. This is separate from Oriel’s optional AI analysis features, and does not mean your research data is shared with the AI tools used during development.