Everything the platform does
The full capability set, ordered by how much it matters rather than by when it arrived.
49 capabilities, live today.
Sorted by importance, not date. Tier 01 is what makes this a platform rather than an assistant — remove any one of those and it stops being the product. Tier 02 is what you touch on an ordinary working day. Tier 03 is the control, governance and scale that make it safe to leave running.
01The core
The mechanisms the whole product rests on. These are the reason work gets done rather than merely assisted.
Describe the work; get a plan you can approve
Brief a recurring piece of work the way you would brief a colleague. A planner turns it into a validated milestone plan — with concrete success criteria, scope boundaries and a cost estimate — that you read, edit and approve before anything runs.
JobsSince March 2026Start from what you already do
Not everyone arrives knowing what to automate. Describe a process you run today in your own words and an interview digs out what makes it executable — the decision rules, the exception paths, the people involved. What comes back is a proposed job with every step marked automated, gated on your approval, or left manual, grounded in the systems, people and knowledge sources you have actually connected. It ends on a live job, not a draft: you decide how it starts — on a schedule, from an intake form created for you, from an email or webhook, or by hand — and every approval you asked for in the interview is carried into the plan as a gate the planner cannot drop.
JobsSince August 2026The executable plan improves itself
Every run ends in a structured review that grades what the job learned and applies safe changes to the executable plan as a new versioned revision. Provenance is complete, every change is reversible, and a circuit breaker stops automatic revision after three consecutive changes fail to produce a passing run.
JobsSince June 2026Approval gates the engine enforces
Consequential and irreversible steps are declared as gates in the validated plan and enforced by the execution engine — not left to a model remembering to ask. Nothing you did not authorise reaches the outside world, and an unanswered irreversible gate fails closed.
ApprovalsSince April 2026An independent verdict that binds
A separate judgment-grade model decides whether the run passed or failed against the objective, external receipts and produced files. The verdict controls the run's outcome; it is not an observability score, and the agents that did the work cannot mark their own homework.
JobsSince June 2026Decide from wherever you are
A gate reaches you on WhatsApp, in the app, by email or on a phone call, carrying enough context to decide safely in seconds. You answer in your own words and the platform reads what you meant. When a plan knows who owns a decision, the question goes to that named person on their channel rather than to everyone.
ApprovalsSince March 2026Human milestones connect agent work to your process
A plan models work only a person can do as first-class human milestones, with an owner, deadline and completion evidence. Agent milestones and human work share one plan, so a job coordinates AI execution with the process your people already run instead of stopping at an approval gate. When that work takes hours or days — a supplier reply, a meeting to agree — the job waits for it durably and releases the dependent steps only when it reaches a real outcome.
JobsSince August 2026Durable execution as the foundation
A job runs server-side for hours, days or weeks, replays after a crash, survives restarts and deploys, and resumes exactly where it stopped. This reliability is table stakes for serious agentic work — the floor beneath the plan, gates and verdict rather than the differentiator by itself.
JobsSince February 2026One process, run for every new record
A scheduled job can retrieve a changing set of records — new emails, new orders, rows that changed since last time — and run the nested process once for every accepted item. The engine does the retrieval deterministically and checkpoints its progress, so a long intake survives retries and restarts, and no model is ever asked to enumerate what should have been fetched.
AutomationSince September 2026An execution of record
Every run leaves a complete account: what was done, in what order, on whose authority, against what evidence, and what it cost. Reasoning traces, tool calls and human responses are all part of the record — and so is a time-ordered ledger of everything the run sent outside itself, each message carrying the provider's own id, so a send can be proved and not merely claimed.
GovernanceSince June 2026A staffed department, not a workbench
A roster of fifteen named specialists — research, planning, data, email, external correspondence, live voice, web reading, documents — orchestrated by a chief of staff. Each has a face and a page; the services you can hand work to are indexed by the question you arrive with rather than by product name. It ships finished, rather than as parts for you to assemble.
AgentsSince February 2026A deliberate mix of models, across vendors
Frontier models from different vendors take the judgment calls; open-weight models take the mechanical, high-volume steps. Each seat is routed to whichever model does that job best for the least money, escalating to a stronger one when a step fails — and re-tuned as the market moves, with nothing for you to choose, learn or migrate.
PlatformSince June 2026Cost attributed to every run
Each model call is metered and attributed to the job, the run and the person who drove it, so a job's price sits next to its result instead of arriving as a monthly surprise.
BillingSince March 2026
02Everyday capability
What you actually touch on a working day — the channels you reach it through, the systems it connects to, and the things it can produce.
Connects to the systems you already run
Google Workspace, Microsoft 365, Dropbox, OneDrive, any mailbox over standard IMAP and SMTP, and Bokio for accounting. Several accounts of the same kind can be connected under distinct names, so agents target the right one explicitly.
ConnectionsSince July 2026Your own internal APIsA doorway a system opens so other software can read and update its data — how the agents talk to your systems directly.
Describe an internal system you built yourself and agents can call it under governed, allowlisted access — credentials held server-side and never visible to the agent.
ConnectionsSince March 2026Triggers: schedule, email, form, SMS, webhook
Start a job on a schedule or from an event outside the platform. Set up in plain language, with a readable URL, a readiness checklist and a way to test against a sample payload before going live.
AutomationSince April 2026Chat on web, WhatsApp and voice
The same conversation reachable from a browser, a WhatsApp thread or a phone call, with session memory that carries across them.
ChannelsSince March 2026Live voice, in the caller's language
Linda answers and places real phone calls, greets each caller in the right language from the first word, and runs saved shortcuts you trigger by saying their name — asking for any detail the shortcut still needs.
VoiceSince August 2026A to-do queue for one-off work
A quick instruction arriving by WhatsApp, phone or email is captured as a lightweight to-do and carried out automatically — no job to set up, no ceremony.
To-dosSince August 2026A persistent workspace per job
Files, a database, a knowledge base and a memory that survive between runs, so a recurring job behaves like a dedicated worker accumulating context rather than a stranger starting fresh each week.
JobsSince June 2026Real files, not approximations
Genuine .docx, .pptx, .xlsx and .pdf output — including data-driven documents with embedded charts — that a recipient can open and use with no conversion step.
OutputSince June 2026Real computation in an isolated sandbox
When a step needs arithmetic rather than estimation, the agent writes and runs real Python in a locked-down sandbox with the data libraries already installed. The script can read your project files, database collections and knowledge corpora directly — so one step can own “read all four hundred and reconcile” — and its reach is fixed by a credential minted for that single run. Binary files are transformed without their contents ever entering the model's context.
JobsSince July 2026A shared document database
Structured data produced by one run is available to the next, editable through familiar spreadsheet vocabulary by people who do not think in schemas.
DataSince June 2026Knowledge corpora you manage yourself
Upload your own material so agents answer from your documents rather than from general knowledge — managed without a developer in the loop.
DataSince February 2026People Coordination
Agree a time with several people outside your company: a professional slot-options email with one-click accept, a response matrix showing who agreed to what, and an operator calendar you can step into. Participants never need an account, and each is written to in their own language.
Agent servicesSince July 2026External correspondence
An agent that holds a real back-and-forth with a supplier, customer or contractor over email — chasing, following up and reporting back — with your project's tone of voice and approval rules applied.
Agent servicesSince April 2026Shared mailboxes, watched and triaged
Point it at a shared inbox over Microsoft 365 or IMAP and it reads every thread, tells a bounce from a customer, sorts mail into categories you define, and keeps one board of what is waiting on you and what is waiting on others — with the overdue ones counted. Arrival rules run in shadow first, reporting what they would have done until you trust them. It drafts suggested replies straight into the mailbox's own drafts folder in your voice, never invents a fact — it leaves a marked gap and tells you what it still needs — and never sends. Message bodies are read and discarded, not stored.
Agent servicesSince September 2026Facility scheduling
Plan a week of slots for teams, rooms or resources: pin the fixed bookings, auto-plan around them, then drag to adjust with push, swap and undo. Share by link or export to xlsx and pdf.
Agent servicesSince July 2026A read-only web operator
When something only exists on a public web page — a price, a listing, a timetable — a strictly read-only browser worker fetches the live page, without opening a door into your own network.
AgentsSince July 2026Research grounded in live search
The research agent draws on live Google results and carries the citations through to the answer, in quick questions and deep research runs alike, so every claim can be checked against its source.
AgentsSince August 2026Dashboards you describe in a sentence
Describe the view you want over a job's real data and get a chartable widget back, or pick from a shelf of sixteen ready-made ones — latest run, reliability, who answers, cost by model, messages by channel. Widgets sit on a board you arrange by hand, and stay where you put them when an agent adds more. Everything runs read-only and network-less, so it can never change or leak what it displays.
ReportingSince June 2026Per-job statistics and time saved
One page per job showing failure rate, cost per run, human time required and the learnings it has accumulated — plus the manual effort each run replaced, compounding over time.
ReportingSince June 2026Watch a run on its own plan, and scrub back through it
A live run is drawn on the plan it is executing: where it stands, what each milestone promised against what it delivered, and whether it is working or waiting on a person. Drag back through the timeline to see the run as it stood at any moment — including the four hours somebody took to answer — and answer an open gate from the graph itself. On a phone the same run reads as a list.
ReportingSince September 2026A value report that says what it measured and what it guessed
One report across the whole account: what the agents did, what it cost, and how much human effort they absorbed — with every figure marked as measured or estimated, in words nobody has to be taught. It answers "is it getting better?" only when there is enough data to support the claim, and cost is shown to account admins only.
ReportingSince September 2026Standing instructions every agent honours
Write your house rules once per project — tone of voice, audience, do's and don'ts, approval rules, glossary — and every agent, in every chat and every job, follows them without being reminded.
GovernanceSince June 2026Memory that curates itself
Durable facts about how you work are remembered and recalled automatically in later sessions, so you stop repeating yourself and never maintain a memory file.
AgentsSince June 2026Notifications kept apart from approvals
Outcomes and progress arrive in a notification feed, acknowledged per person. The approval queue stays exactly one thing: decisions waiting on you.
NotificationsSince August 2026
03Control, governance and scale
What makes it safe to leave running, defensible to an auditor, and sellable to someone else's customers.
Hard spend caps
A monthly ceiling per account, enforced across every channel — chat, job execution, voice, triggers and WhatsApp — so nothing can run up an unbounded bill.
BillingSince July 2026Roles and scoped memberships
People hold explicit roles in each account they can reach, rather than platform-wide access, so least privilege applies and you can audit who can do what where.
GovernanceSince July 2026An audit log of people and agents
A structured, searchable trail of user and agent actions, so incidents, billing anomalies and unexpected changes can be investigated without trawling server logs.
GovernanceSince July 2026Learnings you can read, correct and retire
Every learning a run used is on the page — editable and deletable, with the plan hints it motivated retiring alongside it — so a job's evolution is something you steer, not a black box you trust.
GovernanceSince August 2026Credentials server-side, data isolated
Connection credentials are held server-side and never visible to the agents, which see only allowlisted operations. Data is isolated per account and processed within the EU.
SecuritySince March 2026The white-label service-provider layer
Deliver the whole platform under your own brand: a hierarchy of providers with custom domains, inherited branding, host-scoped sign-in and your own agent naming — no per-customer engineering, and the reseller relationship stays invisible to the tenants below.
Service providersSince July 2026Run your own account
Invite members, reset passwords, see who last signed in, and open the product as one of your users — clearly labelled and time-limited — to reproduce a problem without asking for their password.
AdministrationSince June 2026Drive it from your own tooling
A secure per-account MCPA standard plug (Model Context Protocol) that lets AI tools connect to other software — the same connector everywhere, instead of one custom integration per system. endpoint exposes what a human operator can reach — jobs, chat, approvals, executions, data — so you can run and diagnose the platform from Claude Code or another external runtime without sharing a raw APIA doorway a system opens so other software can read and update its data — how the agents talk to your systems directly. key.
PlatformSince June 2026Sign in with Google or Microsoft
Use the identity your organisation already has, rather than managing another password.
PlatformSince June 2026English and Swedish, at parity
Both languages are first-class across every surface — the app, jobs, live voice, transactional email, approvals and outbound messages — each reader served in their own.
PlatformSince July 2026Usable from a phone
Approvals, job status, results, runs and external conversations all work properly on a narrow screen, because the person a decision is waiting on is rarely at a desk. Menus open as a sheet under your thumb rather than a dropdown built for a cursor, and a running job reads as a list instead of a shrunken graph.
PlatformSince July 2026Built to run under load
Work distributes across the fleet rather than cascading into timeouts: parallel milestones, chat streams that resume across servers, incremental event polling, bounded context per call and spend limits counted fleet-wide.
PerformanceSince August 2026
Want this running in your business?
Join the waitlist and we'll show you what these features look like on your own work.