#Choosing how to build your app's agent
There are three common ways to give your app a conversational agent: build it in your own backend with an agent framework, configure it in a visual builder or a WhatsApp chatbot builder, or have your coding agent build it with Vatio, which runs it. Start with the job the agent has to do. If your app has users, it may need to answer their questions, check their account or act through your existing APIs, including when they write through WhatsApp. Then decide how you want to build, operate and improve the agent once real users talk to it.
#Build it with an agent framework in your own backend
You can ask Claude Code to build the conversation system inside your app, with an agent framework or your model provider's SDK. The framework handles the model calls and the tool loop. You still decide who maintains conversation history, user identity, access controls, human handoff, an inbox, deployments and tests for the mistakes you fix. This fits when you want to own that operation and have someone who can maintain it.
#Use a visual builder or a WhatsApp chatbot builder
If you prefer configuring the agent in a visual editor, or you need to upload contact lists, manage campaigns or design visual flows, evaluate those screens directly in the products you are considering. Check whether each one supports your users' identity, the APIs they need, a person taking over and a way to test updates. Prices and feature lists change: check the provider's current offer and do not rely on an old comparison.
Vatio offers outbound messages through its server API. It does not include a campaign dashboard or a visual flow editor.
#Use Vatio from your coding agent
Vatio is for people building an app with a coding agent (Claude Code, Cursor, Codex) or an app builder (Lovable, Bolt, Replit, v0), live or still in progress. You do not need to write code yourself.
- Step 1, from any chat. Add the remote MCP server
https://vatio.ai/mcpas a connector in Claude, ChatGPT or Cursor, sign in with a Vatio account and ask for the agent. It is created in a preview to try; nothing reaches users until it is published (MCP, Deployments). - Step 2, to build it into the app's code. In the project run
npx @vatio-ai/skills, then/vatio-init(in Cursor or Codex, ask it to use vatio-init), and/vatio-improvedaily.
Your coding agent builds the agent from your app's project. Vatio runs it in your app and, as added channels, on WhatsApp and Instagram, with an inbox for supervision and human takeover. Private tools can use verified user identity; your backend decides which data may be returned. You decide what the agent may do and approve every publication. Vatio has one preview and one live environment per workspace.
After launch, report mistakes in the inbox and run /vatio-improve in your coding agent to prepare fixes and test them against the reported cases. The WhatsApp guide covers testing on the shared preview number and connecting your own number. For prices, see Pricing.
#What to try before choosing
Use a real support question, an action that needs a signed-in user, a case that should go to a person and an incorrect answer you want to fix. Follow each through creation, supervision and a tested update.
#Related guides
- Build an AI agent for your app: Run npx @vatio-ai/skills and /vatio-init in your app's project, answer what the code can't, and try the preview before publishing.
- The AI agent runtime: What Vatio runs between a customer's message and your agent's reply, including grouping messages and pacing.
- An AI agent on WhatsApp: Add WhatsApp to your app's agent: test on a shared number, then connect your business number.
- Add an agent to your Lovable app: Sync the project to GitHub, open it in Claude Code or Cursor and add an agent.
- Your daily improvement routine: Report mistakes in the inbox, /vatio-improve has your coding agent fix them, and you approve the update.