Migração 100% grátis + 1 mês grátis com cupom MIGRAR1MES · novos clientes em planos até R$ 200/mês Migrar agora
Launch offer · R$ 99/month for the first 3 months

Your WhatsApp AI agent, built by our team.

You tell us your goals. Our AI-specialized engineers set up the agent in up to 7 days — CRM Hunter, Sandbox Hermes and Uazapi all configured, with maintenance and 24/7 support included. You just connect your WhatsApp.

  • Setup in up to 7 days
  • Dedicated server included
  • 24/7 maintenance
  • No lock-in
How it works

From the first call to a live agent in 7 days

Our team handles everything. You review and approve.

01

You tell us your goals

A 30-minute call with our team to understand what the agent will do (handle support, qualify leads, schedule, sell) and how it should communicate in your brand voice.

02

Our team builds the stack

We provision the dedicated server, configure CRM Hunter and Sandbox Hermes, connect Uazapi (WhatsApp) and integrate it with your OpenAI API key. We tune the agent and knowledge base.

03

You connect your WhatsApp

You get a WhatsApp number via Uazapi (R$ 38/month, paid directly to Uazapi) and we connect it to the agent. Within 7 days your agent is chatting with your customers.

04

Our team manages it 24/7

Maintenance, updates, monitoring, prompt tuning and human support — all included. You just watch the conversations and the results.

What's included in the monthly fee

No hidden upsells. What you see here is what you get.

Dedicated server for your agent

Exclusive KVM VPS running 24/7. No multi-tenant, no noisy neighbors.

Stack built by our AI team

CRM Hunter, Sandbox Hermes, Uazapi, LLM integration, credential vault and backup — all configured.

Maintenance and 24/7 support

Stack updates, prompt tuning, troubleshooting and human support.

Free daily backup

Workflows, conversations and configurations saved daily. One-click restore.

Pricing

Transparent pricing, no surprises

Everything Rollin Host charges, and everything you pay separately. We show both.

Launch offer

Managed AI Stack

Everything configured by our team

R$ 99 /month

for the first 3 months

Then R$ 249/month · no lock-in · cancel whenever you want

I want my agent
  • Dedicated server for your agent
  • CRM Hunter + Hermes + Uazapi configured
  • Agent training + knowledge base (RAG)
  • Maintenance and updates 24/7
  • Human support
  • Free daily backup
  • 30-day satisfaction guarantee
Use cases

Who benefits from an AI agent

We configure the agent for your specific scenario — no one-size-fits-all packages.

Customer support

Answers FAQs, checks order status, opens tickets and escalates to a human when needed.

Lead qualification

Asks qualification questions, books meetings and hands off the ready-to-close lead to your sales team.

Scheduling (clinics, salons, real estate)

Checks availability, books appointments, sends 24h and 2h reminders, reschedules when needed.

E-commerce and WooCommerce

Answers shipping questions, recommends products, generates payment links, tracks post-purchase orders.

Collections and renegotiation

Reminds about installments, offers negotiation options within your rules, escalates complex cases to a human.

SaaS onboarding

Guides customers through first use, answers real-time questions, collects feedback and flags churn risk.

Comparison

Managed Stack vs alternatives

Rollin Stack DIY Solution Traditional agency
Setup 7 days · done by us 2–8 weeks · you do it 2–4 weeks · they do it
Typical monthly cost R$ 99 (promo) / R$ 249 R$ 800+ (server + time) R$ 2,000–5,000
Maintenance & updates Included 24/7 You handle it Included (but $$)
Human support 24/7 N/A Business hours
Management panel included Yes · CRM Hunter You build it Usually no
LLM cost (OpenAI) You pay directly · transparent You pay directly Bundled (no clarity)
Tech stack

The tools powering your agent

Open-source wherever possible. You are not locked in — take the agent with you if you want.

CRM Hunter (painel) Incluso Hermes (agente IA) Sandbox isolado Uazapi (WhatsApp) OpenAI API · GPT-4 Anthropic Claude (opcional) Google Gemini (opcional) PostgreSQL (memória) Qdrant · Chroma (RAG) Redis (queue) Webhooks públicos Docker + Compose Backup automático TLS 1.3 + AES-256

Panel · CRM Hunter

Configure the agent without code, manage leads, unified inbox, briefings and scheduling — all in one place.

Agent · Sandbox Hermes

Hermes runs in an isolated sandbox · fully configurable via CRM Hunter · self-improving · tool calling support.

WhatsApp · Uazapi

Stable unofficial API · multi-instance · media, audio, lists, buttons · event webhooks. Also available via Meta Official API.

AI · LLM + RAG

OpenAI (default) · Claude and Gemini on demand · Qdrant/Chroma for knowledge base · embeddings.

Frequently asked questions

Can't find your answer? Message us on WhatsApp.

How does the R$ 99 payment work and when does it become R$ 249?

The launch offer is R$ 99/month for the first 3 consecutive months from the date of purchase. From the 4th month onward, the price becomes R$ 249/month. You are notified before the change and can cancel at any time if you are not satisfied.

Why do I need to pay Uazapi separately?

Uazapi is the API that connects your WhatsApp account to the agent. It charges R$ 38/month per number directly — you pay them. We use Uazapi because it is the most stable way to integrate unofficial WhatsApp with automations. If you prefer the official WhatsApp Business API (Meta Cloud API), we can discuss it on demand — the per-message cost is different.

Why do I need to provide my own OpenAI API key?

Because token usage varies enormously between clients — someone handling 50 conversations/month pays US$ 5; someone handling 5,000/month pays US$ 200+. Bundling it into the monthly fee would be unfair to both sides. You create an account at platform.openai.com, generate an API key and share it with us. You pay only what you use, invoiced directly by OpenAI.

Can I use Claude (Anthropic) or Gemini (Google) instead of OpenAI?

Yes, on demand. If you prefer Claude, Gemini, or running an open-source model (Llama, Mistral) on our LLM Hosting, we configure it during setup. OpenAI is the default due to its maturity in tool calling and function calling, but we swap it on request.

How long does it take to have the agent live?

On average 7 calendar days from the initial briefing. Simple cases (support with a small FAQ) go live in 3–5 days. Complex cases (CRM/ERP integration, RAG with a large knowledge base) may take 10–14 days.

Do you train the agent with my knowledge base?

Yes. You send us documents (PDF, spreadsheet, text), URLs from your website, or export from your CRM. We index everything in a vector database (Qdrant or Chroma) and connect it to the agent via RAG (Retrieval-Augmented Generation). The agent then answers based on your data, not just the general knowledge of GPT.

Can I connect with my CRM, ERP, spreadsheet or Google Calendar?

Yes. The Hermes agent integrates via webhooks and APIs with HubSpot, Salesforce, Pipedrive, RD Station, Bitrix24, Google Sheets, Calendar, Notion, Airtable, Postgres, MySQL and any system with an API. We configure the integrations your agent needs.

What if the agent gives a wrong answer? How do you fix it?

We monitor conversations in real time on the CRM Hunter dashboard and identify error patterns during the first weeks. Prompt, behavior and knowledge base adjustments are included in the monthly fee. If you spot an error, open a ticket and our team fixes it.

Can I have more than one agent?

Yes. Each additional agent (one for support, one for sales, one for collections, for example) is priced on demand. Since the stack is already running on your dedicated server, the cost of adding more agents is proportional — much cheaper than the first one.

Can I cancel whenever I want?

Yes. No lock-in, no penalty. Cancel in the dashboard or by opening a ticket and the service ends at the close of the paid cycle. During the first 30 days we offer a satisfaction guarantee — full refund if you are not happy.

Who owns the conversation data? Do you use it to train AI?

You own it. Conversations are stored in the PostgreSQL database on your dedicated server, encrypted at rest (AES-256). We do not use your data to train any model, and we do not share it with third parties. You can export or delete it at any time. Full LGPD compliance.

What is the difference between this stack and building something in-house?

Cost and speed. Hiring a senior developer + server + tooling + maintenance easily costs R$ 5,000–10,000/month between salary, infrastructure and time. Here you pay R$ 99 for the first 3 months (R$ 249 after), you do not need to hire anyone, and you get 24/7 human support. CRM Hunter and Sandbox Hermes come ready out of the box — you just manage the results.

Pronto pra hospedar seu projeto de IA?

Comece em 5 minutos. Migração gratuita, suporte 24/7 em português e garantia de reembolso em 7 dias.