AI Agents with Dify - Build No-Code AI Agents with Dify

Delve into generative AI for agents, covering artificial intelligence, machine learning, deep learning, neural networks, and llms, while noting transformers, content generation, and the need for human oversight.

Understand what large language models (llms) are—deep learning systems that generate text by predicting the next word—and their limits, like outdated knowledge, and how Reg and agents address them.

Explore rag (retrieval augmented generation) and tooling for application programming interfaces to empower llms to fetch private data, search and retrieve relevant contracts, and execute actions via application programming interfaces.

Discover the difference between agents and workflows in no-code AI automation, highlighting determinism versus latitude, goals, tools, and knowledge bases that guide action.

Choose the cloud version with a free sandbox to set up your DeFi workspace, or self-host on AWS, Azure, or on-prem, then create a free account and begin.

Explore the Dify studio environment, discover five application types, and learn how to edit agents, chat flows, and workflows with knowledge bases and tools, plus tracing, monitoring, and publishing options.

Build a no-code text generator app with Dify by defining a prefix prompt and company description variable, generating slogans, testing prompts, adjusting model and temperature, then publish via API.

Build a no-code chatbot that remembers conversation context and handles back-and-forth dialogue, guided by instructions, variables, and tone settings for engaging responses.

Create your first AI agent in Dify by adding tools like Wikipedia and Yahoo Finance, enabling autonomous research and actions beyond chat, turning a chatbot into a capable agent.

Design a no-code ai workflow by wiring a trigger to LLM blocks and inputs to generate a tagline. Explore the visual workflow interface, block chaining, and publishing for app integration.

Build a chat flow, a deterministic chatbot that controls its steps before replying, integrating tools, memory, and translating to French or performing Wikipedia lookups for no-code AI agents with Dify.

Select the right model for each task in a model-agnostic platform, balancing cost, latency, and context length, and switch between providers like bedrock, OpenAI, Azure AI Studio, Gemini, and anthropic.

Compare multiple models side by side to choose the right model for a task, evaluating outputs and latency and refining prompts to improve results.

Tune model parameters such as temperature and max tokens to control output creativity and length, recognizing deterministic behavior at low temperatures and more varied results at higher temperatures.

Explore how to control model outputs with the text output parameter, switching between text and JSON formats, and enforce a JSON object schema for structured results in no-code AI agents.

Select and compare models for external apps, tailoring each workflow step with suitable LMs while balancing cost, context windows, and multimodal capabilities.

Explore how tools empower llms to take actions—calling APIs, sending emails, and performing searches—using a growing marketplace of tools and the option to build custom, no-code ai agents.

Learn how tools and actions require explicit parameters in workflows, and how no-code AI agents with Dify automatically infer parameters for tasks like Wikipedia search and Ford and Honda comparisons.

Discover multi-tool integration for no-code AI agents with Dify, adding Yahoo Finance and Wikipedia tools to fetch ticker, analytics, and news, then post a summary to Slack.

Create a custom tool in AI agents by connecting an API to fetch crypto prices, define an OpenAI schema, and test, save, and deploy it to channels like Slack.

Create and index a knowledge base for ai agents using retrieval augmented generation and chunking. Import from files, notion, or websites and configure chunking and vector search for retrieval.

Learn to build and query a knowledge base in ai agents using chunked documents, vector search, and no-code chatbot workflows in Dify.

Build no-code ai agents with Dify by mastering chunking strategies and retrieval options: create per-agreement chunks, explore vector, full text search, and hybrid search, and tune top chunks and threshold.

Learn how to integrate knowledge with tools to build no-code ai agents, using knowledge retrieval, workflows, and Slack notifications to answer drone manual questions.

Publish and deploy your AI agent by adding a knowledge base and Slack webhook, then run, share, or embed the app via site or Chrome extension for production-ready access.

Demonstrates embedding a chat bot on a website using an HTML tester and iframe code, previewing in real time and testing queries via a bottom-right widget.

Publish and consume your agent via the publishing API, using the base URL and authorization header with the API key, and test streaming chat with Postman.

Learn to run AI apps in batches with Dify by uploading an Excel template to process multiple inputs, and receive batch results for company descriptions and workflows.

Explore logs and monitoring in ai agents with Dify to trace workflow runs, view input-output details, and track token usage for cost insights.

Export and import DSL files to share agents and workflows, manage workspace access with editing roles, and customize agent names and icons before publishing and running the app.

Celebrate completing the course and earning your certificate, access on-demand videos, Q&A, and support, and explore Copilot Studio, Mind Studio, Power Automate, Lang, Lang Flow, and Zapier.

Access all courses at 80–90% off and join the intelligent worker newsletter for gen AI and no-code updates, plus a platform with live Q&A and all content for one price.