Custom workflow agents and chatbots with knowledge, skills, lead capture, and API access. Built from a conversation in Morocco.
A complete agentic workflow, without the complex setup
Build the behavior, context, integrations, and customer experience in one place, then keep control of every publish.
Agent and chatbot builder
Describe the result in plain language. IMFA Agents creates a working draft with an identity, system prompt, greeting, skills, knowledge setup, and customizable chat experience.
Agentic workflows and API
Connect instructions, context, knowledge, skills, models, tools, and actions. Use an agent-scoped REST API to bring the same agent into your own server-side applications.
Customizable and protected
Match the agent to your brand, from its behavior to its widget design. Built-in prompt-injection defenses protect system instructions and treat retrieved content as untrusted data.
One agent, three clear moves
Describe the outcome, add the context your agent needs, and publish one consistent workflow wherever your audience or application needs it.
- 01
A plain-language brief for a new support agent Describe it.
Tell IMFA Agents what the agent should know, do, and sound like. One plain-language brief is enough for the builder agent to create the first draft.
- 02
A node editor with connected agent settings and a brand color palette Refine & customize.
Shape the workflow with your instructions, knowledge, context, skills, models, and tools. Then customize the conversation and design for your brand.
- 03
One agent deployed to a website, Telegram, WhatsApp, and Discord Deploy anywhere.
Publish to your website, Telegram, WhatsApp, and Discord, or connect the agent to your own server-side application through the REST API.
Your agents, powered by the models you choose
Choose from OpenAI, Claude, Gemini, Kimi, and DeepSeek to power your agents. More providers are coming soon.
- OpenAI
- Claude
- Gemini
- Kimi
- DeepSeek
- More models soon
Simple, clear plans
Start free, move to Pro for more capacity and model control, or choose a Custom plan shaped around your needs.
.webp?tr=w-384,q-60,f-webp)
Standard
For building and testing agents with a light weekly workload.
- Build and publish AI agents
- 5 included credits each week
- Default AI models
Pro
For active agents that need more knowledge, model control, and capacity.
- Everything in Standard
- 100 included credits each month
- Add any website to your agent's knowledge
- Custom AI models
- Non-expiring credit packs
Custom Solutions
Custom solutions designed around your business needs.
- Custom AI agents and software solutions
- Strategy and scope built around your goals
- Tailored workflows and integrations
- Hands-on development, launch, and support
.webp?tr=w-384,q-60,f-webp)
Create agents from the tools you already use
Connect IMFA Agents to Codex, Claude Code, Cursor, or VS Code. Your AI coding agent can create, configure, and publish agents without leaving your workflow.
Answers before you need support
Read these before you write to support. They cover most of what comes up in the first week.
[01]How do I put the agent on my site?
Publish it, copy the embed line, and paste it into your page. The widget carries the design you approved in the builder, and every change you publish afterwards goes live in the same minute.
[02]What can the agent learn from?
PDFs, Word documents, Excel sheets, and plain text. Upload them once and the agent reads everything into its knowledge base. It answers from that material, and when a question falls outside it, it says so rather than improvising.
[03]Which channels does it work on?
The widget on your own site, plus Telegram, WhatsApp, and Discord, each under your own bot identity. Voice is there too: your agent gets a line people can call and talk to. Each channel connects from the builder in about a minute.
[04]What happens to the leads it captures?
Whenever a visitor shares an email or a phone number, the agent saves the contact along with a short summary of what they wanted. Everything lands in the Inbox tab of your dashboard, and you get a notification the moment it happens.
[05]Which models does the agent use?
Start with IMFA Agents' managed models without any setup. For more control, connect your own API key and choose models from OpenAI, Claude, Gemini, Kimi, or DeepSeek.
[06]Can I connect an agent built with IMFA Agents to my application?
Yes. Every agent can have a scoped REST API key for trusted server-side integrations. Your application can send messages, continue conversation sessions, read message history, and retrieve captured leads.
[07]How does IMFA Agents defend against prompt injection?
Every published agent receives server-side protection rules that keep system prompts and tools private. Retrieved documents and website content are marked as untrusted data, so they cannot replace the agent's instructions. These defenses reduce risk, but no AI platform can guarantee complete protection.
