Short answers to the questions you should be asking.
Differentiation, data residency, model providers, security review, and what changes after handover.
Common questions, answered straight.
How is this different from ChatGPT or a copilot.
ChatGPT is a model. A copilot is a chat box. Hubzoid is a hub plus agents, built into how your company decides, not bolted on beside it.
- The hub holds your structured knowledge: policies, calendars, vendors, decision trees.
- The agents act on it three ways: on-demand Tools, conversational Q&A, scheduled Background.
- Every answer is grounded in your numbers, not the open internet.
How do we get started.
Request an Agent Map, the fastest way to see what a hub would do, before any commitment.
- Answer a few questions about your team and the tools you run.
- Back comes a one-page shortlist: the agents that would do the most, what each watches, and the hours it gives back.
- No demo, no sales sequence, a real document.
How does pricing work.
Fixed-price, scoped to your company. Exact numbers come in the proposal, after a short fit call.
- Your first agent + the hub is one price. Each agent after that is its own.
- No per-seat pricing, no open-ended scope.
Where do you work, and which systems do you support.
Anywhere. We deploy inside your own cloud.
- Connects to the systems you already run: major ERPs, accounting platforms, and your sector's tools.
- Meets your team where they already are: WhatsApp, Telegram, Slack, web, plus scheduled digests and dashboards.
Do you replace our existing systems.
No. Hubzoid reads from the systems you already run, ERP, accounting, CRM, drives, and chat, and makes them legible to your agents. We do not rip and replace, and we do not copy your data out. Your core systems stay exactly where they are.
How long does an engagement take.
Engagements are scoped to your company during discovery. The depth scales with your environment, integrations, and security review. We share an exact timeline as part of scoping, not as a marketing claim.
Where does our data live.
In your cloud or VPC. Hubzoid does not maintain a customer-data lake. World models, agent definitions, and runbooks are deployed into your environment. Engineer access is time-bound and revoked at handover.
Which model providers do you support.
Anthropic Claude on the customer's contract by default. OpenAI and Google supported on request. Your data trains no model: not the provider's general models, and nothing at Hubzoid, enforced via provider enterprise terms and zero-retention access.
Is the runtime open source.
Yes. The runtime is the hubzoid package: MIT-licensed, pip install hubzoid, source at github.com/hubzoid/hubzoid. It runs inside your own cloud, so your team can read exactly what executes against your data. The same package powers our engagements. What you audit is what ships.
What happens after handover.
The hub is yours. Your team operates it against the runbook. A quarterly retainer is available for the first quarter, to surface drift early. Cancellable on thirty days notice.
Can we start with one team and expand.
Yes. Start with one team or a pair, commonly marketing plus finance or ops plus accounts, and add more once the first agents have earned it. Adding teams later is a follow-on engagement.
Which languages do the agents speak.
Whatever your team uses. English, Hindi, French, German, and other major languages are supported on Telegram, Slack, and digests. Background reports follow the language leadership reads. Tool outputs follow your brand-kit language defaults.
Do you sign InfoSec questionnaires.
Yes. Five business day turnaround target. We share our control documentation under NDA on request.
Are you SOC 2 certified.
Pre-audit. No SOC 2 report yet, and we say so.
- Controls built against the SOC 2 Common Criteria: access control, change management, audit logging, incident response.
- Type I targeted for end of 2026, Type II about twelve months later.
- Meanwhile the hub runs in your cloud, under your keys, every action in your logs.