It depends on your biggest pain point right now:

• Missed calls? → Atlas (AI Receptionist)
• Slow lead reply? → Echo (Speed-to-Lead)
• Few reviews? → Rhea (Reputation Builder, beta July)
• Dormant CRM? → Nova (AI Closer, beta August)
• Marketing autopilot? → Orion (Growth Engine, beta September)

Still unsure? Book a free 15-min discovery call. We'll recommend the agent that pays for itself fastest based on your numbers — and tell you honestly if none of them fit.

DIY (Self-Install): You buy a tutorial (€29-49), follow the video, deploy in 1-4 hours. You manage your own infrastructure (n8n, Twilio, ElevenLabs accounts). Best if you're technical and have time.

DFY (Done-For-You): We deploy for you. €299-999 setup + €99-299/mo subscription. Live in 7-14 days. We handle hosting, infrastructure, monitoring and support. Best if you want results, not work.

About 70% of customers choose DFY for their first agent. After they see it work, around 30% come back to DIY for a second one.

From signing the order:

• DIY tutorials: 1-4 hours of your time, no calendar wait
• DFY Standard (€299): 7 business days
• DFY Pro (€499): 14 business days
• DFY Custom (€999+): Scoped per project, typically 3-6 weeks

The bottleneck is usually getting access to your CRM, calendar and phone system — not our work. Customers who pre-share credentials are live faster.

For a Standard DFY setup, we'll ask for: your business name, current call/lead handling process, sample of recordings or transcripts (for tone), 2 minutes of audio if you want voice cloning, and access to two systems (typically a CRM and a calendar). Total time on your side: about 90 minutes split across the first week.

Yes. On every agent page (Our Agents) you can watch a real demo video showing the agent at work end-to-end. For a live test, book a 15-min call — we'll plug you into a sandbox version of Atlas or Echo so you can call it yourself and judge the experience.

Yes. We're based in France and most customers are in Europe, but we've deployed agents in the UAE, Singapore, Vietnam, Canada and the US. Voice cloning supports 28+ languages. Time zones are not a problem — async deployment works fine, and our support is responsive 9am-9pm CET on weekdays.

You provide 2 minutes of clean audio (a recorded business voicemail works great). We process it through ElevenLabs Voice Lab to create a custom voice model that the agent uses for outbound calls. The result is indistinguishable from your real voice for most callers.

Important: we require a written consent agreement before cloning. The voice is locked to your business — we can't reuse or share it. You can delete it anytime from your dashboard.

Yes — within reasonable boundaries. Our agents (powered by Claude Sonnet 4.6) handle multi-turn conversations, qualification trees, scheduling logic, objection handling and escalation. They know when to hand off to a human, and they don't hallucinate fake information about your business — we ground every conversation in your knowledge base.

What they don't do well: legal advice, medical advice, multi-hour open-ended consultations. For those, we hand off to you immediately.

For text channels (SMS, WhatsApp, email): essentially every language Claude supports — 30+, including French, English, Spanish, German, Italian, Portuguese, Dutch, Arabic, Mandarin, Japanese.

For voice (Atlas), we currently support 28 languages with native-quality output: all major European languages, Arabic, Hindi, Mandarin, Japanese, Korean, Vietnamese, Thai, Indonesian. If your language isn't on the list, ask — we may have added it since this page was updated.

Yes, extensively. During DFY setup we tune: tone (formal/casual/friendly), pace, vocabulary (what industry terms to use, what to avoid), greeting and sign-off, escalation triggers, and edge cases (what to do if the caller gets aggressive, asks for a discount, etc.). DIY tutorial buyers get template prompts they can adapt themselves.

Atlas: Voice (inbound + outbound), SMS, WhatsApp, web chat. Books to Google Calendar.

Echo: SMS, WhatsApp, email, LinkedIn, Instagram, web forms. Pushes to your CRM.

Rhea: Email, SMS, WhatsApp, web. Posts to Google Reviews.

Nova: Email, SMS, WhatsApp, Messenger. Lives inside your CRM.

Orion: Email, social posts, Notion reports. Connects to Meta Ads, Google Ads.

Yes — and it's where the real magic happens. The Growth and Scale subscription tiers are built for this. Common combinations: Atlas + Echo (catches every customer entry point), Atlas + Rhea (answers calls + asks for reviews afterwards), or all 5 together for high-volume operations.

Agents share context (a lead Echo qualifies can be handed to Atlas to call, then to Rhea to review-request). That's the whole point of a marketplace — they're built to work together.

Three layers of fallback:

• Soft escalation: the agent says "let me get someone who can help" and routes the message to your team via SMS/email/Slack.
• Hard escalation: for emergencies (medical, legal, anger), the agent immediately patches the call to a human and notifies you.
• Logging: every failed interaction is logged with full transcript so we can refine the agent monthly.

The setup fee (€299-999) covers one-time deployment work: discovery call, voice cloning, agent configuration, integrations, end-to-end testing and team training. The subscription pays for ongoing operation: hosting, infrastructure, API costs, monthly tuning, support, updates and 24/7 monitoring. If your agent breaks at 3am, the subscription is what ensures someone's on call.

Yes. No long-term contracts, no cancellation fees, no negotiation. You can pause or cancel from your Stripe Customer Portal in two clicks, anytime. We bill monthly in advance, so the worst case is one extra month if you cancel right after billing day. We'd rather lose a customer than trap one.

If after 30 days of being live, your agent isn't producing measurable results — fewer missed calls, faster lead replies, more reviews — we refund 100% of your setup fee. No questions, no friction. You keep the work we did. We'd rather lose €299 than have an unhappy customer trash us in reviews.

This guarantee applies to all DFY setups, including Standard, Pro and Custom tiers.

Solo (500 calls/messages per month) and Growth (2,000 per month) include automatic overage at €0.15 per call or message. We notify you at 80% usage so you can adjust if needed. Most customers never reach the limit — but if you regularly do, upgrading to the next tier is cheaper than paying overage. Scale tier is unlimited.

All payments are processed through Stripe. We accept: Visa, Mastercard, American Express, Apple Pay, Google Pay, and SEPA Direct Debit (for European customers). Annual prepay (10% discount) is available on request — email billing@botlib.ai.

Yes — and it's a common path. About 30% of tutorial buyers come back for DFY after they realize they'd rather focus on their business than tinker with workflows. We credit the tutorial price (€29-49) against your first month of DFY subscription as a thank-you for trying DIY first.

Yes. Our partner program (see Developers · Track B) is built for agencies reselling our agents to their clients. Three tiers — Bronze, Silver, Gold — based on active client count. Top-tier partners earn 35% recurring commission and get co-marketing budget. Email partners@botlib.ai for details.

For DFY subscriptions: No. We bundle API costs (Claude, ElevenLabs, Twilio passthroughs) into the monthly fee. What you see is what you pay.

For DIY tutorials: You'll pay your own infrastructure costs directly to providers. Realistic monthly cost for a small business: €30-80 (n8n cloud €20, Twilio ~€20-40 for typical call volume, Claude API ~€10-20). We provide cost estimates in every tutorial.

Out of the box: HubSpot, Pipedrive, Salesforce, Zoho, Notion CRM, Airtable, Monday, Folk, Attio. Email/Calendar: Gmail, Google Calendar, Outlook, Microsoft 365. Phone: Twilio, Aircall, RingCentral. Messaging: WhatsApp Business, Meta Lead Ads, Instagram, Messenger. If your system isn't listed, we likely support it via Make/Zapier passthrough — ask.

Yes, in two ways. Option A: Twilio number-porting — we transfer your existing number to Twilio (takes 2-4 weeks). Option B: Forward your existing number to a Twilio number — instant, but you keep paying your current carrier. Most customers go with Option B for the first 90 days, then port if it works.

Our Pro and Custom DFY tiers handle custom integrations. We've built one-offs for: ATS systems (Recruitee, Welcome Kit), industry-specific tools (Doctolib, Mindbody, Clinicea), legacy systems via webhooks, and even SAP for one enterprise client. If it has an API or webhook, we can probably connect it.

Yes — every DFY deployment includes a 3-7 day staging phase. The agent runs in shadow mode (sees real conversations but doesn't reply) so you can review every output before flipping the switch. We refine prompts based on what you flag, then go live with confidence.

Our infrastructure has been at 99.93% uptime over the last 12 months. We use multi-region failover (EU-West primary, EU-Central failover) and 24/7 monitoring with automatic alerts. If we miss our 99.5% SLA on Scale tier, you get pro-rated credit automatically — no need to ask.

Continuously. Every month we ship: prompt refinements based on your conversation logs, model upgrades (Claude releases improve agents automatically), new channel support, and bug fixes. DFY customers receive these updates without lifting a finger. DIY tutorial buyers get notified via email when the underlying workflow gets a major upgrade — they can re-import the JSON anytime.

Yes, by design. Your customer data is:

• Hosted in the EU (Frankfurt and Paris regions) on AWS infrastructure
• Encrypted at rest (AES-256) and in transit (TLS 1.3)
• GDPR-compliant with a signed Data Processing Agreement available on request
• Never used to train models — we use Claude API with zero-retention enabled
• Yours, always — full export available anytime, full deletion on cancellation

We're working towards SOC 2 Type II certification (target Q3 2026).

Primary region: AWS Frankfurt (eu-central-1). Failover: AWS Paris (eu-west-3). All conversation logs, customer records and configuration data stay in the EU. Voice clones are stored on ElevenLabs EU infrastructure under our enterprise account. We do not use any US-based subprocessors for storage.

Yes. We're registered with the CNIL (French data protection authority), maintain a public privacy policy, sign DPAs with every customer, support data subject requests (access, portability, deletion) within 7 days, and notify breaches within 72 hours as required. Our DPA is available for download from your customer dashboard or by emailing legal@botlib.ai.

Two scenarios:

• Automated processing: Yes — our pipelines read conversations to generate replies. This is fully encrypted and never seen by humans.
• Human review: Only when you explicitly opt-in for monthly tuning, or when you submit a support ticket referencing a specific conversation. Even then, only the named conversation is reviewed, by a single named engineer, with a logged audit trail.

You can disable human review entirely in account settings if your industry requires it (we have customers in legal and healthcare who do).

You have 30 days from cancellation to export everything (conversation logs, customer records, prompts, voice clone files). After 30 days, data is automatically deleted from production systems and 90 days later from backups. We provide a deletion certificate on request.

Voice calls handled by Atlas are recorded by default for quality and training purposes. The recording disclosure plays at the start of every call ("This call may be recorded for quality"). For text channels, conversations are logged automatically — there's no legal disclosure requirement for written communication in most jurisdictions, but we recommend adding a footer to your forms ("Replies may be reviewed for quality").

You can disable recording entirely in account settings, but it disables monthly tuning.

Apply through the Become a Creator page. Required: a working agent (with at least 1-2 paying customers), a 60-90s demo video, and a clear answer to "what problem does this solve and for whom?". We review every application within 5 business days. If approved, we onboard you over 14 days: production help, listing copy, payment setup, launch.

Depends on the product type and traffic source:

• Tutorials, you bring the customer: You get 80%, BotLib 20%
• Tutorials, BotLib brings the customer: 50/50
• Subscriptions, you bring + maintain: You get 50%
• Subscriptions, you bring only: You get 40%
• Subscriptions, BotLib does everything: You get 15%

Full table on the creator page.

Always. You retain full IP ownership. We license a non-exclusive right to distribute on our marketplace. You can sell on Gumroad, your own site, or anywhere else simultaneously. The only exception is our optional Featured program (6-month exclusivity in exchange for top placement and marketing) — and that's opt-in, not default.

It's a tier-based affiliate program for agencies and tech consultants who deploy our agents for their own clients. Three tiers: Bronze (1-4 clients, 20% recurring), Silver (5-14 clients, 30% recurring), Gold (15+ clients, 35% recurring + co-marketing). Email partners@botlib.ai with your client list to apply.

Monthly payouts via Stripe Connect, Wise or bank transfer. Payouts happen between the 5th and 10th of each month for the previous month's revenue. Minimum payout threshold: €50 (rolls over if not reached). Detailed earnings report with every payout — you see every sale, every refund, every API cost deduction.

Three channels:

• Email (hello@botlib.ai): replied within 4 hours on weekdays, 24h on weekends
• In-app chat (Growth + Scale subscribers): typical response under 1 hour during business hours
• Phone (Scale tier only): direct line to your dedicated success manager

For genuine emergencies (agent producing wrong info, broken integration affecting business), email with "URGENT" in the subject — we page someone immediately, 24/7.

Yes. Solo tier includes 1 admin user. Growth includes up to 5 users with role-based permissions (admin / editor / viewer). Scale is unlimited users with SSO support (SAML, Google Workspace, Microsoft Azure AD). User management is in your account settings.

Upgrade anytime — change takes effect immediately, billing is pro-rated. Downgrade anytime — change takes effect at the end of the current billing cycle. Both are self-service from your Stripe Customer Portal. If you need help, email billing@botlib.ai.

Try the password reset link first. If that doesn't work, email hello@botlib.ai from the same address you signed up with — we'll verify and unblock within an hour during business hours. If your account uses SSO and the SSO is broken, we can issue a temporary login link to bypass.