We’re building native experiences on shared information, identity and authority. Bring people, AI and systems into the work with the context they need and the control you choose.
A note can grow into a project. A conversation can carry a task. A capture can carry an agreement about who controls it. Thread the information together for useful AI, with deliberate rights over how it is used.
These are the connected experiences we’re building. Identity and communication serve the whole environment as well as their own management and conversation surfaces. Each story below explains the job, the mechanism and what stays under your control.
01 / Identity & authority
Know who is acting. Decide what they can do.
Contacts are one view of a deeper Identity Graph connecting people, organizations, devices, AI agents, services and places. Roles, credentials and delegated authority give those connections meaning.
Your control
Choose which identity to present, what facts to establish and which details or actions to disclose and delegate in each context.
One communication foundation for conversations, AI handoffs, system events and commands. Typed messages connect participants, source objects, task state and authority.
Your control
Choose participants, hosting and delivery routes, plus the rights attached to shared content and each requested action.
A library of connected objects, with original content, versions, provenance and relationships. Files, collections and search offer different ways into the same information.
Your control
Decide where content is held and how it may be used. Track changed rights and removal requests across known participating holders.
Attach an evolving, authored note thread to a person, place, file or task. Keep observations, revisions and source context connected as a thought develops into a decision or work.
Your control
Keep a note private or choose its shared audience. Connecting it to an object does not make it public.
Start with a to-do. Grow into the way your team works.
Compose tasks, projects, phases and sprints from shared items, relationships and rules. Boards, timelines and dependency views reveal the same work at different levels of detail.
Your control
Delegate a defined contribution with the context and action permissions it needs.
From evidence to report. With the reasoning intact.
Build reports, workbooks and briefings from structured claims, calculations and source evidence. Keep the displayed result linked to the information and revision that support it.
Your control
Scope AI edits to the relevant material, and choose live references or a fixed published revision.
Connect geography, infrastructure, observations and changing conditions in a native model. Bring authorized personal or operational context into route and asset decisions.
Your control
Keep private searches, routes and observations on infrastructure you or your organization control.
Connect photos, video and audio to their subjects, source components and rights. Find the meaningful moment, preserve its context and assign control through accepted agreements.
Your control
Assign technical control to another person or device, delegate capture rights for a defined window, and specify the audience and permitted uses.
Navigate purpose, requirements, interfaces, callers, dependencies and exact source at progressive depth. Retain useful understanding across tasks instead of repeatedly rediscovering it.
Your control
Scope each person or AI’s context and change authority to the task.
The 8OS vision / Connected context, programmable rights
Connect more. Disclose less. Keep control.
Useful AI needs the relationships between your information: the people, conversations, evidence and purpose. We’re building those connections together with the rights that govern their use and cryptographic evidence for the facts that matter.
AN AGREEMENT BEFORE THE CAMERA ROLLS“Capture on your device. Put these recordings under my control.”
01 / AGREE
One hour. Defined captures.
Both participants accept the terms. New captures in the agreed window are assigned to the designated controller.
02 / USE
One task. Selected material.
The controller permits an assistant to prepare a transcript from chosen clips, with an explicit processing destination.
03 / ACCOUNT
Know what was permitted.
Keep the agreement, source clips, derived transcript and recorded actions connected. Track expiry, changed rights and managed-copy removal.
Illustrative product direction: technical control and permitted use, with signed acceptance. This does not establish legal ownership of the recording.
Rights that describe the actual use
Our smart-contract direction connects accepted agreements to specific people, devices, information and actions. Define what can be read, edited, shared, delegated or supplied to AI, for which purpose and for how long. Make a transfer of control an explicit, accepted event.
Prove the fact. Keep the detail private.
Our zero-knowledge work targets proofs of defined conditions without revealing the private information behind them. For example, prove possession of an eligible entitlement without disclosing its secret identifier. Signed agreements establish who accepted terms; ZK proofs establish the particular condition their proof system checks.
Make withdrawal an accountable operation
Connect rights changes to the managed information they affect. Keep revoking access, deleting data and destroying keys distinct. Record which participating holders completed a requested operation, which remain pending and what each receipt or proof actually establishes.
Give AI the thread it needs
Let an authorized assistant follow a requirement into a conversation, a document, a recording or code. Scope its context to the task, preserve the sources behind its result, and keep permission to read separate from permission to send information to another AI service or act on it.
What does cryptographic evidence establish?
A signature establishes acceptance by a signing key. A ZK proof establishes a defined statement against particular inputs. An erasure receipt records a scoped operation by a known holder. Verification also needs the right identity, authority and current state. None of these proves total security or the removal of unknown exports, screenshots or information learned by an external model.
Identity & authority / Shared service + management surface
An identity for every participant. A purpose for every permission.
We’re building Identity around the real relationships behind a name: who someone represents, which device is acting, what a credential establishes and what an assistant has permission to do. A contact list, device roster and service directory become useful views of the same foundation.
Manage the relationship behind the contact: who represents whom, which claims support that relationship and what each participant may do.
Keep contexts distinct
A person can represent a company, contribute to a project and participate privately. Shared names or contact details should never silently combine those contexts or grant access.
Give authority a precise scope
Bind an AI or service mandate to particular information, actions, recipients, time and budget. Keep identifying a participant separate from authorizing its next action.
Establish a fact without sharing everything
Use selective disclosure and, for suitable claims, zero-knowledge proofs to establish an entitlement without exposing its private details. Connect the proof to its trusted issuer, purpose and current authorization state.
Carry agreements and rights with the relationship
Connect accepted terms, licences and service entitlements to the people, organizations and devices they cover. Keep a visible history of who granted a right, what it permits and when it changes.
01Participant
02Context & credentials
03Scoped authority
One participant can hold different roles and mandates in different contexts. Product direction.
Picture this
Bring a specialist into a project, let them present a relevant credential, and give their assistant a bounded review task. The project can recognize their contribution without absorbing their unrelated private identity or information.
A conversation for you. A coordination layer for everything that acts.
The messaging vision spans human-to-human, human-to-AI, AI-to-AI and system communication. People need readable conversations. Agents need precise task handoffs. Systems need typed events, commands and outcomes. 8OS connects those needs through shared identities, information and controls.
Move from conversation to coordinated action without reconstructing the participants, evidence and task context at every handoff.
Share the object with its rights
Link the relevant file, source version, place or observation into an exchange. Connect its context and permitted uses to the conversation, including who may receive, forward or supply it to AI. Changes to those rights belong in the same history.
Give agents continuity
Carry the goal, permitted actions, deadline, progress and unresolved questions with a task. Let an agent request clarification, return a result or hand off work without a human chat window.
Know what actually happened
Keep message delivery, task execution and independent verification distinct. A received command is not a completed action; the outcome belongs in the record.
Give communication a home you choose
Organize spaces and conversations on chosen infrastructure. Design local exchanges to continue through an internet outage, with visible policies for queued or paused external delivery.
01Event or conversation
02Scoped task & decision
03Action & recorded outcome
Illustrative coordination across a person, an AI agent and a system. Product direction.
Picture this
An equipment monitor reports an anomaly. An authorized agent gathers evidence and proposes a repair. A person accepts the specific action, and a service reports the result. The readable discussion and precise machine record describe the same sequence.
We’re designing Storage to preserve more than a file’s bytes. Its source, history, relationships and permitted uses matter wherever it appears. Native apps work with that connected object; familiar file interfaces provide a bridge to the tools you already use.
Carry the relationships that make information useful, alongside familiar files and folders. Reuse the same source across your work without losing track of which version supports it.
Give one source many useful views
Bring the same object into a folder, project collection or search result. Keep its identity, versions and relationships across those views instead of creating a separate library for every app.
Choose its home and manage its lifecycle
Place selected collections on your phone, workstation or chosen mesh nodes. Make offline availability visible. Track requests to revoke access or remove managed data, including holders that have not yet completed them.
Keep familiar tools in the picture
Use native file interfaces to reach connected content. Make external editing explicit, retain the original and reconcile a returned file as a new revision.
Take useful context with you
Preserve source history and relationships in supported structured transfers, alongside original-file export. Let your choice of AI work from an authorized view of the library.
01Original & history
02Collections & app views
03Chosen devices & tools
One connected information model, with deliberate storage and disclosure choices. Product direction.
Picture this
Keep a field report in your library, reference it in a project and share a permitted view in a conversation. If you open a copy in an external editor, make that handoff and the returned revision visible.
Notes is the quick way into the connected environment. We’re building an experience where you can capture an observation beside the thing it concerns, return to its history and let it grow into useful work.
A note accumulates understanding around its subject. You can see who added an observation, what changed and how it connects to the next decision.
Capture in context
Start a note on a person, organization, place, device, file or work item. Keep the subject attached as you add entries over time.
Keep the history of understanding
Retain authored, dated and versioned entries. Distinguish a new observation from an edit to an earlier one, with a path back to its source.
Let the idea become work
Connect an observation to a decision or task. The shared work model lets the detail grow without severing the note that explained why it mattered.
Choose when to involve others
Keep personal annotations private or work in a deliberately shared thread. Search and AI should receive only the note context that their task is permitted to use.
01Subject & observation
02Authored note history
03Decision or connected task
An annotation stream builds understanding around its subject. Product direction.
Picture this
Record a concern beside an inspection photo, add a later observation and turn the issue into a task. The reviewer can follow the task back through the note history to the original evidence.
We’re designing Projects so a small task can grow into a coordinated programme without losing its history or meaning. Templates make the first step familiar. Underneath, shared items, relationships and rules let a team shape how work is organized, assigned and accepted.
The same contribution can belong to a sprint, a customer commitment and a wider programme. Different views help each participant act while keeping the dependencies and completion criteria connected.
Shape the process around the work
Start from a checklist, project or phase-gate template. Add fields, relationships and rules as the job grows; reuse a useful structure as your own template.
Zoom from purpose to the blocker
Move between objectives, requirements, tasks and dependencies. Read a board, timeline or dependency graph over the same work, then follow the detail into a document, conversation or code change.
Make completion mean something
Define which evidence satisfies a task and which conditions permit a phase to close. Derive progress from the work beneath it, using the completion rules the team has chosen.
Coordinate people, AI and systems
Give each contributor a role, relevant context and bounded authority through Identity. Connect human tasks, agent work and system jobs to their dependencies and recorded outcomes.
01Note or to-do
02Tasks, collections & rules
03Evidence-backed progress
One work model, from an individual contribution to a coordinated programme. Product direction.
Picture this
A customer request begins as a note and a to-do. As it grows, organize the work into phases, assign people and agents, and inspect the blocker on a timeline. Close the milestone when its required evidence is present, with the original request still connected.
Our document research starts with a knowledge object. Original evidence, structured claims, calculations and document layout remain connected but distinct. A report, workbook or briefing becomes a purposeful view of shared information, with a traceable path back to its sources.
Reuse a supported conclusion across a report, workbook and presentation, while retaining the evidence and revision that make it trustworthy.
Follow a claim to its source
Connect a statement or value to the original page, cell, observation or media region, together with units, assumptions and the source revision.
See what a change affects
Trace new evidence into dependent calculations and conclusions. Distinguish live references from fixed accepted revisions so a published report remains reproducible.
Give AI the relevant part
Use the document’s structure and dependencies to focus an edit on a bounded section. Preserve the surrounding definitions and references without repeatedly supplying the whole document.
Write, calculate and present
Create a report, work through a calculation and assemble a briefing from approved sections, charts, maps and media. Familiar authoring surfaces share the underlying content and its evidence.
01Source evidence
02Claim & calculation
03Report, workbook & briefing
Different presentations share meaning, with explicit revisions and provenance. Product direction.
Picture this
Turn an inspection observation into a sourced calculation, then reuse the accepted result in a report and briefing. When the observation changes, identify what needs review while retaining the previously approved packet.
A living model of places, journeys and infrastructure.
We’re building toward maps that connect public geography with your permitted private context. Geometry, network connections, typed restrictions, time and evidence each have a role. The map is a view into that model, serving everyday journeys and deeper infrastructure questions.
Treat the map as an interactive view of connected places and infrastructure. Your private observations and constraints give the geography practical meaning.
Plan for the actual journey
Account for accessibility needs, vehicle configuration, changing loads and arrival conditions. Separate requirements that must be met from preferences among feasible alternatives.
Understand dependencies and change
Connect an asset to its inspections, services and restrictions. Ask what a change affects and inspect the evidence, time and uncertainty behind the answer.
Operate on your terms
Keep rendering, search, routing and necessary AI within chosen infrastructure. Prepare regions for disconnected use and make missing or stale coverage visible.
Work with the object on the map
Select the place or asset itself, then inspect its attributes, observations and connections. The native view is designed to query and render the geography as you move, with private context attached to the relevant object.
01Place & infrastructure
02Private context & conditions
03Choice with supporting evidence
A place is connected to the people, systems and decisions that depend on it. Product direction.
Picture this
Plan a delivery with a changing load. A bridge restriction may matter on one leg and not the next. Connect the route choice to the applicable restriction, its source and the time it applies.
We’re building Media around both meaning and control. A capture connects what it depicts to its source, components and governing agreement. The person holding the camera can agree that another participant controls the new recordings, with that authority represented from the moment of capture.
A connected media object carries more than pixels: its source, subjects, edits, relationships and rights. Those connections make it useful to people and AI while giving each use an explicit scope.
Assign control before capture
Use an accepted, time-bounded agreement so captures made during a defined session are governed by another person or device. Bind the recording to that agreement at capture time, without a separate transfer for every photo.
Transfer rights by agreement
Make a later transfer of technical control a signed offer and acceptance. Identify the new controller and any access the sender retains. Smart contracts express the participants, material, terms and permitted actions.
Find the meaningful moment
Combine subject, place and time with transcripts, frames and semantic search. Expand from a scene overview into its participants or a precise recording segment as the question requires.
Preserve the path to the original
Keep edits, annotations and extracted observations linked to the source capture. Distinguish original evidence from an interpretation or derived result.
Share a deliberate view
Develop selective disclosure around regions, layers and components, with separate treatment of the original and derived versions. The audience and policy are part of the sharing decision.
Compose edits from their sources
Represent a sequence through source segments and edit operations, with audio, depth and other components connected to the capture. Keep the edit recipe and original distinct from a flattened export.
01Accepted capture agreement
02Media, components & rights
03Scoped human & AI use
The design assigns technical control and permitted uses. Legal ownership remains a separate question. Product direction.
Picture this
Agree to film someone on your device with their control over the next hour’s captures. They permit an assistant to transcribe selected clips for a project. The agreement, original recordings and transcript remain connected, with the processing destination and subsequent rights changes visible.
Zoom from the reason for a product to the line that serves it.
CodeGraph connects software structure with the wider information that explains it. The aim is more useful understanding per token: find existing capabilities, explore their implications and supply the depth a task needs.
Give each person or AI the depth their task needs, from the product decision to the affected source, and preserve that understanding for the next contribution.
Find the capability before rebuilding it
Use structural and semantic relationships to discover existing behavior even when the search terms differ from the code’s names.
Understand the consequences of a change
Follow callers, dependencies, fan-in and fan-out across repositories. Connect those paths to requirements, interfaces and the evidence for changing them.
Carry context across the work
Move between a concise overview and precise source. Keep decisions and verified understanding available to the next person or AI contribution.
Make a contribution within its scope
Connect a proposed change to its requirements, affected consumers and validation. Let reviewers and agents work within defined access and change authority while keeping the wider purpose visible.
01Product purpose
02Capabilities & dependencies
03Exact source & change
Zoom in for precision. Zoom out for purpose and consequences. Product direction.
Picture this
Start with a user requirement, locate the capability that already serves it, inspect the affected callers and dependencies, then return to the product context to judge the proposed change.
Apps give information its purpose. Infinite Reasoning Fabric is the shared service being designed to connect goals, evidence and useful understanding across their work.
Identity, permissions, mesh connectivity and reasoning belong beneath the app experience. A participant should get the context their task needs and the authority to make their contribution.
Built around a domain. Connected by a wider vision.
BioTwin, SovClear and Digital-Fabric are full application projects with their own domains and identities. We keep their ownership and integration arrangements distinct from the native 8OS app family.
APPLICATION PROJECT
BioTwin
A domain application connecting scientific information, models and evidence. BioTwin consumers already inform the exploration of useful native operations in 8DB.
Planned to run on 8OS. This direction is separate from current runtime qualification and ownership.
Counterparty intelligence for ownership and sanctions reviews. Connect historical relationships, source records and evidence gaps to an assessment a reviewer can explain.
Explore the project’s current pilot offer and evidence requirements.
A programmable data and settlement service being built on 8DB. Connect digital rights and value with the application data and controls their transactions need.
Node licensing is its first proving ground. Development and qualification follow the project’s own scope.
Listing a project here does not imply 8Braid ownership or a completed 8OS integration. Explore each project’s own product and engagement information.
Build on the right layer
Your application. A stronger foundation.
Embed 8DB in a product you build, or explore an application’s place in the wider 8OS environment. Start with the job, the information and the control your users need.
