GenAI Overview

Generative AI is at the heart of modern intelligent applications. Flow-Like provides everything you need to build sophisticated AI-powered workflows—from simple chatbots to complex multi-agent systems with knowledge retrieval.

No Code Required

All GenAI features in Flow-Like are designed for visual builders. You don’t need to write code to create powerful AI applications!

What Can You Build?

With Flow-Like’s GenAI capabilities, you can create:

Application Type	Description
Chatbots & Assistants	Conversational AI with memory and context
Knowledge Bases (RAG)	AI that answers questions from your documents
Data Extraction	Automatically pull structured data from text
AI Agents	Autonomous assistants that use tools and make decisions
Content Generators	Create text, summaries, and creative content

Core Concepts

Before diving in, here are the key concepts you’ll work with:

1. AI Models & Providers

AI models are the “brains” behind your GenAI applications. Flow-Like supports dozens of providers including OpenAI, Anthropic, Google, and local models via Ollama.

👉 Learn about AI Models & Setup

2. Chat & Conversations

Build interactive chat experiences with full conversation history, streaming responses, and the Chat UI.

👉 Learn about Chat & Conversations

3. RAG & Knowledge Bases

Give your AI access to your own documents using Retrieval-Augmented Generation (RAG). Upload files, create embeddings, and search semantically.

👉 Learn about RAG & Knowledge Bases

4. AI Agents

Create autonomous AI agents that can use tools, search databases, call APIs, and make multi-step decisions.

👉 Learn about AI Agents

5. Extraction & Structured Output

Use AI to extract structured data from unstructured text—perfect for forms, document processing, and data pipelines.

👉 Learn about Extraction & Structured Output

Quick Example: A Simple Chatbot

Here’s a high-level view of what a basic chatbot looks like in Flow-Like:

┌─────────────────┐     ┌──────────────┐     ┌─────────────────┐
│   Chat Event    │────▶│  Invoke LLM  │────▶│  Push Response  │
│  (receives msg) │     │  (generate)  │     │  (send reply)   │
└─────────────────┘     └──────────────┘     └─────────────────┘

1. Chat Event – Receives the user’s message and chat history
2. Invoke LLM – Sends the conversation to an AI model
3. Push Response – Streams the AI’s response back to the user

You’ll learn how to build this and much more in the following guides!

Choosing the Right Approach

If you want to…	Start here
Build a conversational assistant	Chat & Conversations
Answer questions from your documents	RAG & Knowledge Bases
Extract data from PDFs or emails	Extraction & Structured Output
Create an autonomous agent with tools	AI Agents
Use a local/self-hosted model	AI Models & Setup

Prerequisites

Before building GenAI apps, make sure you have:

1. Flow-Like Desktop installed (Download)
2. API keys for your chosen AI provider(s), or a local model running via Ollama
3. Models configured in Flow-Like (AI Models Setup)

Model Availability

Some features like vision (image understanding) or tool use require specific models. Check your provider’s documentation for supported capabilities.

Next Steps

Ready to start building? Choose your path:

• New to AI? Start with AI Models & Setup to configure your first model
• Building a chatbot? Jump to Chat & Conversations
• Have documents to search? Head to RAG & Knowledge Bases