Frequently Asked Questions

Getting Started

Q: How long does it take to create an agent?

A: About 5 minutes. Fill in name, role, instructions, and you're ready to test.

Q: Do I need technical skills?

A: No. You write instructions in plain English. The agent understands natural language.

Q: Can I test before going live?

A: Yes. Use Human in the Loop mode to review all responses before they're sent.

Q: What's a good first agent to build?

A: Start with customer support. It's straightforward and has clear escalation rules. Once you're confident, build more specialized agents.

Q: How do I get my agent's email address?

A: When you create an agent, we automatically generate an email address (e.g., support-bot@customagents.io). You can customize it or use your own domain.

---

Configuration

Q: What's the difference between autonomy modes?

A:

• Human in the Loop: You review and approve every response
• Semi-Auto: Agent sends after 2 minutes (you can cancel)
• Full Auto: Agent sends immediately

Start with Human in the Loop. Move to Semi-Auto after 50-100 emails.

Q: How do I write good instructions?

A: Be specific. Include examples. Define boundaries. Mention tone and style.

Good: "Answer questions about pricing, features, and billing. Escalate technical issues."
Bad: "Be helpful and answer questions."

Q: Can I change settings later?

A: Yes. You can update instructions, tone, autonomy mode, and escalation rules anytime. Changes take effect immediately.

Q: What's the difference between tone options?

A:

• Professional: Formal, business-like
• Friendly: Warm, approachable
• Casual: Relaxed, informal
• Formal: Very structured, official
• Empathetic: Compassionate, understanding

Can I have multiple agents?

Yes! Add as many agents as you need — pricing is a flat $20/month for unlimited agents, plus usage. You can manage them all from a single dashboard.

Can agents learn from conversations?

Currently, agents don't learn from individual conversations. However, you can update your knowledge base based on common questions or issues you notice.

What AI model does CustomAgents use?

We don't disclose the specific model. You don't need to know — just give your agent instructions in plain language and it handles the rest.

Knowledge Base Questions

What file formats do you support?

We support:

• PDF
• Word Documents (.docx)
• Plain text (.txt)
• Markdown (.md)
• Web page URLs (we fetch and extract the linked page — this reads a single page, it does not crawl a whole site)
• Google Docs (via link)

How large can my knowledge base be?

Each document can be up to 20 MB. You can have unlimited documents in a knowledge base.

Can I update my knowledge base without retraining?

Yes! Knowledge bases are dynamic. Changes take effect immediately (within a few seconds). No retraining needed.

How do agents search my knowledge base?

When an agent needs context, it retrieves the most relevant stored content by matching the words in the incoming message against your documents. Write your documents using the same terms your customers use, and keep entries specific — that produces the best matches.

Integration Questions

Which integrations do you support?

We currently have a dozen live integrations — connected with one click:

• Instagram (posts, reels, comments, DMs)
• LinkedIn (text, image, and video posts)
• Google Drive (view, create, and share documents)
• Shopify (products, orders, and customer management)
• Eventbrite (create and manage events)
• Canvas LMS (classroom support for professors and students)
• Outlook (calendar events)
• Calendly (scheduling)
• Stripe (payment links)
• Canva (design assets)
• Luma (events)
• Meetup (events and groups)

Channels (email, text, Slack, in-app) are separate from integrations — they're built in and need no external connection.

More integrations are in progress. Request a new integration.

Are integrations secure?

Yes! We use OAuth 2.0 for all integrations. Your agents never see your passwords or API keys. We only request the minimum required permissions.

Can I revoke an integration?

Yes, you can disconnect any integration anytime from your agent's settings. The agent will immediately lose access.

Autonomy Mode Questions

What's the difference between the autonomy modes?

• Human in the Loop: Agent drafts responses; you must approve before sending.
• Semi-Auto: Agent queues responses with a delay window (2-5 minutes) for review.
• Full Auto: Agent sends responses immediately without human review.

Can I change autonomy modes?

Yes, you can change the autonomy mode anytime. Changes take effect immediately for new incoming emails.

What happens if I'm not available to review in Semi-Auto mode?

If the delay window expires and you haven't reviewed the email, it sends automatically.

Is Full Auto mode safe?

Full Auto mode is safe if your agent is well-trained with a comprehensive knowledge base and clear instructions. We recommend starting in Human in the Loop or Semi-Auto mode and moving to Full Auto once you're confident.

Escalation Questions

What happens when an email is escalated?

When an email is escalated, the agent holds its reply for review and notifies you (the agent's admin): a push notification, an entry on the Escalations view in the dashboard, and — if you've turned it on — a text to your admin number. The alert includes the sender, subject, and the reason it was flagged.

Can I customize escalation rules?

Yes! You can set up rules based on:

• Tone (angry customers)
• Keywords (specific phrases)
• Categories (unrecognized intents)

What's a typical escalation rate?

Most customers see 5-15% escalation rates. If you're seeing higher rates, your agent might need better training or clearer instructions.

Technical Questions

Do you have an API?

Yes! We provide a REST API for:

• Creating and managing agents
• Sending emails programmatically
• Retrieving message history
• Managing integrations
• Webhook subscriptions

View API Documentation (coming soon).

Do you support webhooks?

Yes! You can subscribe to events like:

• New incoming email
• Agent response sent
• Email escalated
• Integration connected/disconnected

What's your uptime SLA?

We aim for 99.9% uptime. We monitor our systems 24/7 and have automated failover in place.

How long do you keep message history?

We keep all message history indefinitely (unless you delete it). You can export your data anytime.

Privacy & Security Questions

Is my data encrypted?

Yes! All data is encrypted in transit (TLS) and at rest (AES-256). We also encrypt sensitive fields like API keys and OAuth tokens.

Do you sell my data?

No, we never sell or share your data. We only use it to provide the CustomAgents service.

Can I delete my data?

Yes, you can delete individual messages, agents, or your entire account. Deletion is permanent.

Are you GDPR compliant?

Yes, we comply with GDPR, CCPA, and other privacy regulations. We have a Data Processing Agreement available for enterprise customers.

Billing Questions

When am I charged?

You're charged on the same day each month (the day you signed up). If you cancel, you won't be charged again.

Can I add or remove agents anytime?

Yes. Billing is a flat $20/month for unlimited agents, so you can add or remove agents whenever you want at no extra charge — changes take effect immediately.

What happens if I run out of usage credits?

Usage (AI tokens, SMS, web lookups) is paid from your credit balance. If the balance runs out, your agents pause processing new messages until you top up. You can turn on auto-recharge so they never pause.

Do you offer annual billing?

Yes! Annual billing is available at a 20% discount. Contact sales for details.

Still Have Questions?

• Email us at support@customagents.io
• Check our documentation
• Join our community Slack