EXTENDED COURSE

Dify

LLM-powered apps to build with Dify — chatbots, RAG assistants that answer over your own documents, and multi-step agent workflows — each assembled on a visual drag-and-drop canvas rather than written in code. Dify is open-source and self-hostable, with a Service API (its own chat-messages / completion-messages endpoints) so every app you build here is callable from your own code.

1Lessons5step-by-step, ~90 min each

RAG chatbots & knowledge apps

Agents & visual workflows

2Cheat sheetcopy-ready expressions

3Examples5what people built

Customer & client portals1

Dify Founder Small biz

In the gallery

Customer support agent over product docs

An agent that answers incoming support questions by retrieving relevant sections from product documentation, then composes a natural-language reply — escalating to a human when confidence is low.

Routine FAQ traffic is deflected automatically, cutting first-response time from hours to seconds.

Try it yourself

Import product docs → Dify Knowledge base → Create Agent app (tool: knowledge retrieval) → Set fallback: "I'll connect you with the team" when score < threshold → Connect via API to your support widget.

Internal tools & ops1

Dify AI-native SME

Course starter

Internal policy and HR Q&A bot

Indexes company handbooks, HR policies, and onboarding documents into a private Dify knowledge base, then serves a chat interface where employees get instant, document-grounded answers to policy questions.

HR ticket volume for common policy questions drops, and answers are traceable to the exact policy clause.

Try it yourself

Upload HR docs (PDF/Word) → Dify Knowledge base → Chatbot app (private, SSO via Dify's built-in auth) → Embed in company intranet as iframe.

Content & marketing1

Dify Creator AI-native SME

Course starter

AI content brief generator from brand guidelines

An agent that takes a keyword or topic, retrieves the brand voice guidelines from a knowledge base, and outputs a structured content brief with angle, audience, key points, and suggested headline options.

Writers start every piece with a consistent, on-brand brief instead of a blank page, cutting brief-writing time by around 80%.

Try it yourself

Upload brand guidelines doc → Dify Knowledge base → Agent app → Workflow: user inputs keyword → retrieve brand rules → LLM generates brief (JSON: angle, audience, key points, headlines) → display formatted output.

Research & data tools1

Dify Scientist

Course starter

Literature review assistant over uploaded papers

Researchers upload a batch of PDFs from a systematic review, then ask cross-cutting questions ("which studies used a randomised design?") and get synthesised answers with paragraph-level citations.

Early-stage literature synthesis that takes days in a spreadsheet is turned into an interactive Q&A session.

Try it yourself

Upload paper PDFs → Dify Knowledge base (split by paragraph, embed) → Chatbot or Agent app → System prompt: "Answer from the provided literature only; cite paper title and page."

Knowledge & docs1

Dify Scientist

Course starter

Lab protocol Q&A chatbot

A RAG chatbot built on the lab's own protocol PDFs and SOPs. Researchers ask questions in natural language and get precise, cited answers drawn from the internal document corpus.

New lab members find protocol answers in seconds instead of hunting through shared drives or asking a senior colleague.

Try it yourself

Upload PDFs (SOPs, protocols) → Dify Knowledge base (chunk + embed) → Create Chatbot app → Configure retrieval (top-k, similarity threshold) → Embed on internal wiki.