RAQ.COM — BUSINESS WORKSPACE
More than 70 business and AI tools under one account — a browser interface for people, an agent interface for systems such as Codex and Claude Code, and a shared record of everything either does.
The shared record
ONE ACCOUNT · PEOPLE + AGENTS · LIVE
01 — PLATFORM
Every separate subscription is another login, another bill, another set of permissions and another audit trail. Raq.com folds the jobs into one account — and gives agents a common route into all of them.
Small and medium-sized businesses usually build their operating system by accumulation: a chat product, a task board, a signing service, a file-sharing account, a form builder, a monitoring service and several AI subscriptions. Each arrives with its own login, billing, permissions, data and audit history — and an AI agent becomes less useful when it can see only one slice of that setup.
Raq.com puts those jobs in one account and gives agents a common route into them. A task can start with a message, use a file from File Drop, create a document, request approval, update a board and leave the result where the team can inspect it. That continuity between tools is what the product is built around.
Browser interface for people
Write messages, manage boards, upload files, review generated media, approve requests, edit documents and inspect reports — the normal interfaces for a team at work.
Agent interface via MCP
Compatible agents get structured operations — list the boards, create a card, upload a file, run an image batch, read a project record. No imitated clicks where a direct operation exists.
A shared operational record
Role-based access and audit logging show who did what. A person can see and amend work created by an agent; an agent can continue work started by a person.
An honest note
Raq.com does not claim feature-for-feature equivalence with every mature specialist product. Its public comparisons describe overlapping workflows, not exact replacements. The value is the breadth of useful tools, the shared account and the agent layer across them — keep specialist systems where you need them and connect them through APIs, Automations or a Custom Module.
02 — THE CATALOGUE
The public catalogue covers the work companies usually spread across separate services. Select a job to see what does it.
The day-to-day record of the business: channels, boards, project hubs, files and the measures that keep a team aligned.
Leads and pipeline, branded client portals, shared support inboxes — and the campaign machinery around them.
Draft, compare, sign and share — with audit records, expiring links and passkey-protected credentials.
Send one question to several models, compare the answers, run a deliberation and keep the reasoning on the record.
Batch generation and finishing across image, video, music and speech — with human review built in.
Build and host sites, watch uptime and page changes, check search readiness and query company records.
A small set for life as well as work: compliance, spreadsheets, meals and health.
For agent operations rather than ordinary chat: execution, supervision, approvals, hand-offs and the record left behind.
03 — WORKFLOWS
A single tool can be useful on its own. Because the records are shared, work can move across the account without a new integration for every pair of products. Four patterns the product is built around:
A
A prospect completes a form. An automation creates or updates the CRM record, opens an onboarding card and creates a client portal. E-Sign sends the agreement; the signed document and client uploads land in the account’s files. An agent can check the whole chain and prepare anything missing.
B
The team stores its rules in Brand Guidelines and starts from a campaign brief. Marketing Campaigns prepares the plan; Rewriter or Ghostwriter produces the copy; Image Lab and Video Lab produce the media; Creative Review collects the decisions; Social Scheduler prepares channel-specific versions and publishes after approval.
C
Website Build & Host creates or updates the site. Page Feedback collects pinned review comments. The SEO & AEO Checker reviews the result, while Site Uptime and Page Watch keep watch on availability and content changes. Help Desk and Website Chatbox AI take incoming support, and a task or automation connects each alert to the responsible person.
D
A person records the desired outcome in Priorities or on a Task Board. Orchestrator starts the relevant agent workflow; Nature Work or Agent Operator HQ shows the live sessions. Ask My Agent requests a human decision when the job reaches a controlled boundary, and the final files and history land where the team can inspect them.
04 — AGENTS
Many business products bolt a chatbot onto an existing app. Raq.com exposes the operational tools themselves, so agents take structured actions against the account — while the browser stays open for people to supervise and correct.
BUILT IN
Describe a job in plain English and Raq.com’s built-in agent selects suitable platform tools and runs it. It works with the same account, files and team as any connected external agent — an agent route for non-technical users, with no coding environment to configure.
VIA MCP
Codex, Claude Code or another compatible client connects through the Model Context Protocol and works against the same records as the browser.
→ tools/list
→ tools/call · list_task_boards
→ tools/call · create_card { board: "site-refresh" }
→ tools/call · upload_file { name: "brief-v2.pdf" }
→ tools/call · run_image_batch { n: 8 }
→ tools/call · read_project_record { id: "PB-114" }ORCHESTRATED
Automations run multi-step workflows across the tools — react to an event, move data, ask AI to prepare material, send work for approval and continue after the decision. Orchestrator, Nature Work, Agent Operator HQ and Priorities cover longer jobs: execution, supervision, hand-offs and the record left behind.
Approvals by design
Sensitive or external actions can use explicit approvals. Ask My Agent runs structured person–agent exchanges with two-stage approval controls; Voice Calls is approval-controlled. Access is account-scoped and tool access can be limited. Work stays visible in the product — a shared record, not activity hidden inside a chat or a terminal.
05 — EXTEND
Extend the platform in two clearly separated ways. The distinction matters: the builder, the runtime and the responsibility are different.
SELF-BUILT
MANAGED
The monthly module fee covers design, development, hosting, maintenance and changes requested during the subscription. Larger jobs are identified during the mock-up stage.
06 — PRICING
Per-user subscriptions. The whole public catalogue is included in each paid plan; higher plans add AI usage and higher service limits.
£14per user / month
The full catalogue at entry limits. Website Build & Host available as a paid per-site add-on.
£45per user / month
More AI usage and higher limits. Voice Calls starts here.
£200per user / month
More AI usage again, with higher service limits across the account.
£500per user / month
Maximum AI usage and service limits.
Prices exclude VAT. Upgrade at once or schedule a downgrade for the next billing date; the total changes as users join or leave. AI work draws on the account’s monthly AI allowance — if it runs out, AI tools pause until renewal or a plan change, and non-AI tools continue. Monitoring tools and FinUK have plan-specific quantity limits. Custom Modules sit outside the plan price at £190 / $259 per module per month.
07 — SECURITY
Raq.com separates how it handles core text AI from image, video and voice providers — and publishes the difference.
Core text and chat requests use Zero Data Retention routes by default: under those routes, the model provider cannot keep prompts or responses for training. You can change the AI Privacy setting to reach models that do not support this protection, and live web search has its own setting, because a current-information lookup may need a standard provider route.
Image and video tools send the prompt and required generation parameters to a specialist provider; voice and transcription tools send what the job needs, and completed media is downloaded to Raq.com’s own storage. Those providers operate under their own API retention terms — the text-AI Zero Data Retention statement should not be applied to every media tool.
Raq.com does not train models on customer content.
08 — COMPANY
Raq.com is built by Vu, an AI and software company based in Solihull, United Kingdom. Vu uses Raq.com for its own work and builds the managed Custom Modules.
PEOPLE
Phil Webb
CEO & co-founder
Alex Hawke
CTO & co-founder — leads product & engineering
OFFICE
Suite 15, 1666 High Street
Knowle, Solihull
B93 0LY, United Kingdom
Free trial — no card required. One use of each AI tool to try.