Everything in the box. No tiers, no add-ons, no modules to buy.
42 shipped · 5 groupsDelivery
10Projects, boards, critical-path scheduling, time, and executable SOPs.
Switch between 'This organisation' and 'All organisations' with live KPIs, task queues, rearrangeable widgets, and monthly agent token meters.
Combines the agency tasks assigned to you with private personal projects and notes that nobody else in the organisation can see, including managers.
Drag-and-drop @dnd-kit boards, 8 semantic project colors, sub-projects, sub-tasks, task crew co-assignees, keyboard card movement, and recurring task loops.
Task dependencies with automated critical-path calculation and cascading date shift previews (/api/gantt/schedule).
Start a live timer straight from any task card, view the full-screen flip clock, or log weekly billable and non-billable timesheets.
Month, week, and day time-grids with toggleable layers for Meetings, Tasks, Content, and Events, plus 1-click task creation from meeting notes.
9 node types (Trigger, Step, AI Task, Decision, Automation, Review, Wait, Condition, Finish), immutable versioning, bottleneck health view, and n8n flow import.
Attach formal specification and QA review documents (.review-doc) directly to tasks and pull requests with sign-off checklists.
Dedicated tabs on every project for creative files, reference links, hour/fee budgets, and client-facing board visibility.
Dock navigation to the left rail, right rail, top bar, or floating bottom dock with instant layout persistence.
Collaborate
7Channels, voice notes, permission-scoped chat AI, block docs, and search.
Public and private channels, direct messages, nested thread replies, @mentions, file attachments, and custom status indicators.
Record voice notes in-channel with a live input level meter and play back with interactive scrubbable bar waveforms.
Ask @td about open tasks, due dates, or workload inside any thread. Context is strictly intersected with what all channel participants can already access.
Rich block documents with slash commands, task checklists, tables, code blocks, inline AI editing, and @ resource tagging for files and folders.
Jump to any task by reference code (e.g. KITE-142), project, note, client, or team member across all of your organisations.
Live workload view, role & project access matrix (Manager, Admin, Member), and per-member hourly cost and billable rates.
In-app notification bell, desktop alerts, and batched HTML emails with 1-click unsubscribe that never fire for bot messages.
Business
10Retainers, DB-enforced scope & approval gates, portals, intake, vault, and CMS.
Give clients their own login (account_type = 'client') to view shared tasks, comment, chat directly with your team, and download invoices.
A task shared with a client cannot be closed until both the agency lead and the client have signed it off — in the app, through the API, or by an agent.
Flag an out-of-scope request on the board and it cannot be closed until a priced change order is approved, billed, or deliberately written off by an admin.
Clients submit structured support tokens with checklists from their portal that convert directly into tracked cards on your project board.
Track monthly retainer hour caps, roll over or bill overages, attach approved change orders, and generate itemized PDF invoices.
Drag-and-drop public forms that automatically spin up a project, file an onboarding task, provision portal access, or encrypt credentials into the vault.
Team and personal credentials stored encrypted (AES-256), revealed for 30 seconds on click, every reveal logged, and never visible to an AI agent.
Custom content collections with your own fields, revisions, scheduled publishing, a media library, a read-only delivery API, and instant hosted client sites.
Plan and review social, blog, and campaign deliverables across every client brand in a unified calendar.
Pooled margin accounting sum(num)/sum(den), overservicing radar, estimate accuracy, and post-publish content traffic/conversion telemetry.
Agents
9Multi-agent sessions, 3-tier verification harness, brand kits, and studio flows.
Lead agent breaks briefs into tasks with explicit acceptance criteria, delegates to specialist agents, and streams live tool transcripts.
Tier 1 deterministic checks (no placeholders, verified URLs from tool transcript), Tier 2 executable checks, and Tier 3 grounded LLM judge.
Any external publish or destructive tool call pauses execution and displays the exact stored JSON arguments for human Approve or Deny.
Researches live web sources via Exa, drafts long-form articles in the client's Brand Kit voice, verifies every citation link, and publishes via connector.
Upload CSV or XLSX files and ask questions in plain English; writes and runs real Python over your numbers and outputs reproducible code.
Transforms client briefs into branded multi-tier proposals with hour estimates and print-ready PDF layouts.
Maintains per-client logos, color palettes, typography, and voice rules automatically injected into every agent run.
Project asset folders with drag-and-drop upload, in-app DOCX/PDF/video inspection, and bulk ZIP export.
Create custom agents with scoped tool permissions, persistent project_memory facts, and metered organisation token budgets.
Code
6Isolated git worktrees, self-healing builds, human PR review, and in-app IDE.
Every @Claude task checks out a fresh git worktree on branch claude/<task-code> off default branch so your working tree is never touched.
Runs tsc and npm run build inside the worktree; if compilation fails, pipes compiler stderr back to Claude to self-heal up to 2 times before pushing.
Caches project_repo_map, enforces that the agent can only write files it first read, blocks empty/unchanged commits, and protects lockfiles/.env/CI.
Compares human pull requests against the linked Bizdaz task brief, leaving line-anchored review notes on GitHub where brief and diff disagree.
Self-hosted Monaco side-by-side and inline diff viewer synced with Light, Dark, and Black themes, modified file tree, and AST check panel.
HMAC-SHA256 signed Next.js/WordPress connectors (300s replay window) that publish entries and pull nightly GA4/GSC metrics to auto-file refresh tasks.
One workspace, one set of permissions.
Every rule below holds everywhere, not just in the screens — the mobile app, your automations, anything you build on the API and the AI agents are all held to it in exactly the same way.
| Can | Manager | Admin | Member | Client portal |
|---|---|---|---|---|
| See every project in the organisation | ● | ● | Led or granted | — |
| Create, rename or delete a project | ● | ● | — | — |
| Internal channels, discussions and team notes | ● | ● | ● | — |
| Client-visible tasks and shared review documents | ● | ● | ● | ● |
| The client's own chat thread | ● | ● | ● | Theirs |
| Retainers, budgets and invoice lines | ● | ● | — | Theirs |
| Shared team items in the password vault | ● | ● | ● | — |
| Somebody else's personal vault item | — | — | — | — |
| Agent runs, agent budgets and token meters | ● | ● | — | — |
| Change somebody's role | ● | — | — | — |
| Transfer ownership of the organisation | Owner | — | — | — |
You only ever see one workspace
Everything you look at is already limited to the organisation you are in, before it reaches the screen. Switching workspace switches what exists for you — there is no setting to get wrong and no view that quietly includes another client's work.
Clients are a different kind of account
A client login is not a cheap seat on your team. It is a different kind of account with its own portal, and your internal channels, team notes and unshared tasks are not reachable from it at all — not hidden, not greyed out, not there.
The vault, stated plainly
Passwords are stored encrypted, revealed for thirty seconds when you click, and every reveal is logged with your name against it. Vault contents never go out through the API, a webhook, or into an AI agent's prompt. But we hold the key that unlocks them in order to show them to you, which means this is strong encryption at rest — not the kind where nobody but you could ever read them. If you need the second thing, we are not it.
What shipped recently
- 10 October 2026Talk to your clients, and work their tickets on your board
The two things that used to happen in email now happen where the work is: a conversation with each client, and their support tokens sitting on the project they belong to.
- 27 September 2026We're now Bizdaz
Same product, same team, new name. Teamdeck is Bizdaz, and it lives at bizdaz.com.
- 26 September 2026Voice notes, a chat assistant, and search that spans every organisation
A batch of things that make the day-to-day faster: say it instead of typing it, ask the workspace a question, and find anything from anywhere.
- 18 September 2026v1.4.0Email notifications
Bizdaz now writes to you when something needs you, so work doesn't wait for the next time you open the app.
Direct answers
How does the Code Agent run without messing up our developers' active branches?+
When you mention @Claude on a task, it takes its own private copy of your default branch and works there, on a branch named after the task. It never touches the branch anybody is working on. It runs your own type-check and build commands against that copy, fixes what it broke up to twice, and only pushes and opens a pull request once the build is green. Merging stays a human clicking the button.
What happens when a client asks for out-of-scope work in the portal or chat?+
You toggle 'Out of scope' on the task card. From that moment the task simply cannot be closed — not by a team member in a hurry, not by an automation, not by an agent — until a priced change order has been approved by the client, attached to an invoice, or deliberately written off by an admin. It is not a reminder or a warning banner; the work stays open until somebody decides who is paying for it.
Can clients see our internal chat, hourly cost rates, or private tasks?+
No. A client login is a different kind of account from a team seat. It can reach the tasks you have marked client-visible, its own chat thread with you, review documents you have shared, its retainer and its released invoices — and nothing else. Your internal channels, what your team costs per hour, the password vault and every unshared task are not hidden from that account, they are unreachable by it.
Is the password vault end-to-end encrypted?+
Plainly: no. Vault items are stored encrypted with AES-256, shown for thirty seconds when you click to reveal, and every reveal is logged with a name against it. They never leave through the public API, a webhook, or into an AI agent's prompt. But we hold the key that unlocks them, because that is what lets us show them to you in your browser — so this is strong encryption at rest, not the kind where nobody but you could ever read them. If your clients require the second kind, use a dedicated vault for those credentials.
Can I run more than one agency or studio under a single login?+
Yes. One account can belong to multiple organisations and create up to five workspaces. The sidebar switcher changes your active organisation in one click, and the Dashboard's 'All organisations' mode aggregates every open task, review gate, and unread mention across all of them.
What does 'Free to start • the first person to sign up runs the organisation' mean?+
You can create your workspace immediately without a credit card or sales call. The first person to sign up becomes the organisation Manager with full control over projects, roles, client portals, and agent budgets, and can invite team members and clients from day one.
Open the studio at seven. Everything is already in sync.
Free to start • the first person to sign up runs the organisation