Still have questions?
We're happy to talk through anything that isn't covered here before you commit to anything.
Organized by where you probably are in your decision. If something isn't here, ask us directly.
The questions people have when they're still figuring out if this is right for them.
An AI agent is a piece of software that can take actions — not just answer questions. It can search the web, draft and send emails, read and write files, call APIs, and work through multi-step tasks without you manually doing each step.
A "team" is a set of agents, each configured for a specific role. Instead of one general-purpose AI tool, you have a Researcher, a Writer, a Client Manager — each trained on how your business works and given the tools relevant to their job.
ChatGPT and Claude are conversational AI interfaces — you ask, they answer. They don't take actions, don't have memory of your business, and don't run without you actively prompting them.
Coworkforce agents are trained on your specific business context, can take real actions in your tools, run scheduled tasks without you initiating them, and remember what they've learned across sessions. They're closer to software you deploy than a chat window you use.
We designed Coworkforce for owner-operated businesses where the person running the business also does most of the work — consultants, advisors, small agencies, e-commerce operators, local service businesses, and small professional services firms.
We also have templates for technical teams (software development, project management, customer success) for small companies that want to extend their team's capacity without hiring.
If you're a solo or small-team business doing repetitive, structured work that benefits from automation — research, writing, communications, scheduling, reporting — this is probably a fit.
Yes. The platform itself has a web interface your agents run inside — you don't manage servers or write code. We handle provisioning, configuration, and maintenance.
The onboarding process is designed around a business conversation, not a technical setup. We ask how your business works. You don't need to understand how AI agents work to use them effectively.
What happens after you sign up, and what we need from you.
Your team is typically live within one week of your intake call. That includes provisioning your dedicated server, configuring your agents, and running test tasks before anything goes live.
The intake call itself is usually 60–90 minutes. The more prepared you are with examples of past work and context about your business, the faster your team can be configured.
The more context the better. Things that are genuinely useful: examples of past work you're proud of (emails, proposals, reports), a sense of your typical client or customer, the workflows that take you the most time, and any tools you use regularly (CRM, email, calendar, file storage).
You don't need to have this organized. The intake process is designed to surface it through conversation.
Yes. Coworkforce is BYOK — you bring your own API key from a model provider (Anthropic, OpenAI, or any OpenAI-compatible endpoint). Your key lives on your server. We never see it, and we never mark up API costs.
Getting an API key typically takes a few minutes. We'll help you through it during onboarding if you haven't done it before. API costs are usually a few dollars per month for light use.
Yes, and this is expected. Agent configurations evolve as your team learns your business. You can adjust agent instructions, add context, change tool access, and add or remove agents through the platform interface.
Major reconfigurations (adding a new type of agent, significantly changing workflow scope) are handled in your monthly tuning sessions. Small adjustments you can make yourself.
What day-to-day use actually looks like.
Through the Open WebUI interface on your workspace (e.g., yourbusiness.coworkforce.ai). It's a chat-style interface where you can send tasks to individual agents or to your team. Agents can be set up to run on schedules or triggered by specific events.
Most day-to-day use is natural language: "Draft a follow-up to the Smith proposal." "Prepare a market overview for tomorrow's call." The agents know enough about your business to interpret context you don't explicitly state.
You correct it. Every action is logged, so you can see exactly what happened. Corrections you give are how your team gets better — provide specific feedback about what was wrong and what you wanted instead.
For high-stakes workflows (sending emails on your behalf, modifying real data), we recommend reviewing outputs before enabling autonomous execution, at least until you're confident in how an agent handles your context.
Yes, if you've configured them to. Agents can run scheduled tasks (daily reports, weekly summaries, follow-up sequences) without you triggering them manually. Your server is always on.
Whether they take real actions while you're offline depends on what you've authorized. You decide what happens automatically vs. what waits for your review.
Coworkforce agents support standard tool integrations through the skill library: web search, file operations, calendar access, email, and a growing set of third-party integrations. Custom integrations are available on Scale and Custom plans.
What tools your specific agents can access is configured per-agent during setup. You can restrict or expand tool access at any time.
The questions that matter most if you're handling sensitive client or business data.
On your dedicated VM. We provision a server for you — no shared infrastructure, no cross-tenant data paths. Your documents, your agent outputs, your conversation history, and your configurations all live on that server.
Coworkforce's infrastructure is used for provisioning and account management. We do not have persistent access to your VM's data after setup.
No. Because you bring your own API key, your agents' LLM calls go directly from your server to the model provider — not through Coworkforce. We never see your prompts, your responses, or the content your agents are working with.
We collect anonymized usage metrics (sessions, errors, feature usage) for operational purposes. We do not collect the content of your agent conversations.
No. Your data is not used to train Coworkforce models — we don't train models. It is also not available to your model provider for training purposes, as long as you use API access rather than the provider's consumer products (which have their own terms).
Most enterprise API plans (OpenAI, Anthropic) explicitly exclude API calls from training data by default. We recommend reviewing your model provider's terms to confirm this for your specific plan.
Your server is deprovisioned after a 30-day wind-down period. During that period, you can export your data — documents, agent configurations, and conversation history. We'll send you instructions for how to do this when you notify us of cancellation.
After deprovisioning, your data is deleted from our infrastructure. We do not retain backups of customer data after the wind-down period ends.
We're happy to talk through anything that isn't covered here before you commit to anything.