Organise your AI work. Prove your process. Protect what was said.
Group AI conversations into named collections — by matter, engagement, project, or theme. The organisational structure is auditable. The conversation content stays encrypted with your key.
A flat chat list breaks down when work gets real
Most AI tools give you a chronological list of conversations with auto-generated titles. That works for casual use. It does not work when:
- You have been working on a client matter for three months across dozens of conversations
- A compliance officer asks you to account for AI usage on a specific engagement
- You need to hand over AI-assisted research to a colleague joining your project
- A regulator asks whether AI was used in producing a specific deliverable — and how
The problem is not that your conversations are disorganised. The problem is that you cannot organise them without creating a record, and you cannot create a record without exposing the content.
Every other AI tool forces you to choose between findability and confidentiality.
CloakAI eliminates that choice.
Everything you need to organise AI work
Structure that is visible when it needs to be. Content that is encrypted because it must be.
Named Collections
Create organisational units — call them matters, engagements, projects, or anything that fits your work.
Move and Organise
Move existing conversations into collections at any time. Build structure as you go.
Encrypted Content, Auditable Structure
Content encrypted with AES-256-GCM. Structure visible for audit. By design.
Timeline View
See when each conversation was created, last accessed, and how it relates to others.
Cross-Collection Search
Search by collection, date range, or conversation title across hundreds of conversations.
Archive and Close
Mark collections as active, completed, or archived. Nothing deleted unless you choose.
Colour-Coded Organisation
Assign colours and icons for visual navigation.
Export Audit Trail
Export structured record of collection organisation. Content stays encrypted.
How it works
Collections add structure without changing how you work.
Create a collection
Name it anything — a matter reference, project code, engagement name, or personal theme.
Assign conversations
Move existing conversations in, or start new ones directly inside a collection.
Work as normal
Same zero-knowledge encryption. Same AI capabilities. Collections add structure without changing how you work.
Review and audit
View the full organisational structure — which conversations belong where, when they were created, and how they relate.
Export the record
Generate a structural export showing your collection organisation. Content stays encrypted throughout.
Who it’s for
Collections work for anyone who needs to account for how they use AI — or simply wants to stay organised.
Legal
Privilege-preserving AI records per matter. Group all conversations related to a case or transaction without exposing what was discussed.
Consulting
Engagement-scoped AI usage. Keep every conversation tied to the right client engagement for governance and handover.
Finance
Deal-level AI governance. Group conversations by transaction, audit, or compliance review with a clear audit trail.
Research
Study-level organisation. Keep literature reviews, methodology discussions, and analysis separated by research project.
Engineering
Feature-scoped AI work. Organise architecture discussions, debugging sessions, and code reviews by feature or sprint.
Healthcare
Clinical governance. Structure AI-assisted research and policy work with auditable organisation that respects patient confidentiality.
Corporate Teams
Department-level accountability. Give every team a way to organise their AI usage and demonstrate responsible use.
Individuals
Personal goal and theme organisation. Group conversations by project, learning goal, or area of interest.
Confidentiality preserved by design. Auditability built into the structure.
| What is auditable | What stays encrypted |
|---|---|
| Collection names and descriptions | Content of every conversation |
| Which conversations belong to which collection | All prompts and AI responses |
| When each collection was created, modified, or archived | Documents referenced in conversation |
| When each conversation was created and last accessed | Everything protected by AES-256-GCM |
This separation is not a policy promise — it is a structural guarantee. The auditable layer exists because it needs to be reviewable. The content layer is encrypted because it must not be. There is no admin override. The architecture enforces the boundary.
How CloakAI compares
Flat list, no grouping, no audit trail, content readable on OpenAI servers.
No user-defined organisation, tied to Microsoft infrastructure.
No project organisation, no encryption, conversations used to improve services.
Copying removes from encrypted context, becomes plaintext elsewhere.
User-defined collections, full audit trail, conversation content encrypted with your key, nothing readable on our servers.
Start organising your AI work today
Create your encrypted vault and start grouping conversations into collections in minutes.