Learn to build event-driven workflows with complex branching, looping logic, and concurrent executions.
Beginner1h19m
Event-Driven Agentic Document Workflows
Instructor: Laurie Voss
Earn an accomplishment with PRO · Updated Oct 7, 2025
- Beginner
- 1h19m
- 8 Video Lessons
- 6 Code Examples
- 1 Graded Assignment PRO
- Earn an accomplishment with PRO
- Updated Oct 7, 2025
- Instructor: Laurie Voss
LlamaIndex- Learn more aboutMembership PRO Plan
What you'll learn
Build an event-driven agentic workflow that parses forms, retrieves relevant information from a source document using RAG, and returns answers to the form’s fields as structured outputs.
Incorporate human-in-the-loop to a document workflow to provide feedback to the agent through text and voice instructions.
Course recap
PROThis course, built in partnership with LlamaIndex, teaches how to build agentic document workflows (ADW) using LlamaIndex's event-driven workflow abstraction. ADWs go beyond simple RAG by applying agent workflows to solve real business problems with structured, specific outputs rather than plain-English answers.
Concept map
Key concepts
- Retrieval-Augmented GenerationEmbedding-based retrieval to give LLMs access to private data within context window limits
- AI AgentsSemi-autonomous software that can be given tools and a goal, working out how to solve problems without explicit step-by-step instructions
- Event-driven workflowsBuilding blocks of agentic systems in LlamaIndex: steps connected by events with branching, looping, parallel execution
- Agentic Document Workflows (ADW)Applying agent workflows to document processing for structured output beyond simple Q&A
- Human-in-the-loopUsing InputRequired and HumanResponse events to pause workflows, collect feedback, and incorporate it into subsequent iterations
- Concurrent executionFiring multiple events in parallel and collecting results with `context.collect_events`
Lesson highlights
- 1.Introduction — Course overview, ADW concept, LlamaIndex workflow abstraction
- 2.What are Agentic Document Workflows — RAG fundamentals, agents, workflows, and ADW concepts
- 3.Building a Workflow — Multi-step workflows, branching, looping, concurrent execution, event streaming
- 4.Adding RAG — LlamaParse for resume parsing, vector store creation, query engine, FunctionCallingAgent
- 5.Form Parsing — Parsing job application forms into fields, concurrent query generation, collecting responses
- 6.Human in the Loop — Feedback events, InputRequired/HumanResponse, LLM-based feedback interpretation
- 7.Use your Voice — Multimodal feedback via Whisper audio transcription, Gradio microphone interface
- 8.Conclusion — Summary
About this course
Learn to build Event-Driven Agentic Document Workflows, in this new course, built in partnership with LlamaIndex, and taught by Laurie Voss, VP of Developer Relations.
Manually processing complex documents can be time-consuming, and error-prone. But what if an AI agent could automate it for you? Agentic document workflows are agent-based applications that you can use to automate an end-to-end document processing workflow. This course teaches you how to build an agentic document workflow, using an event-driven architecture, a design pattern that allows the agent to perform its tasks asynchronously and in parallel.
You’ll design an event-driven agentic workflow that fills out a PDF form based on information from a source document. The agent will use RAG to retrieve relevant data from the source document, parse the form to identify the required fields, convert the blank spaces into questions, and send those questions to the RAG system. You’ll collaborate with the agent using a human-in-the-loop feedback approach through text and with your voice.
In detail, you’ll:
- Understand the basic concepts of agentic document workflow, which consists of RAG, agents, and event-driven workflows.
- Build a series of LlamaIndex’s workflows that increase in complexity and involve branching, looping logic, and concurrent executions.
- Set up the agent’s RAG capability by parsing the source document, loading the extracted information into a vector store, and building a query engine on top of the store.
- Implement workflow steps that enable the agent to parse the form to be filled, turn the parsed information into simple questions, and use the questions to query the RAG pipeline.
- Incorporate human-in-the-loop into the workflow and ask the agent to re-answer when necessary to produce more accurate form responses.
- Add multimodal capability to the agent, allowing it to process spoken feedback.
By the end of this course, you’ll have built an event-driven agentic workflow that fills out a document and responds to human feedback for more accurate form completion.
Who should join?
Anyone who has basic Python knowledge and basic understanding of object-oriented programming, and wants to learn to build event-driven agent workflows!
Course Outline
8 Lessons・6 Code Examples- IntroductionVideo・3m
- What are Agentic Document WorkflowsVideo・5m
- Building a WorkflowVideo with Code Example・18m
- Adding RAGVideo with Code Example・9m
- Form ParsingVideo with Code Example・6m
- Human in the LoopVideo with Code Example・8m
- Use your VoiceVideo with Code Example・6m
- ConclusionVideo・1m
- Quiz
Graded・Quiz
・10m - Appendix – Tips and HelpCode Example・10m
Elevate your learning experience with Pro
Upgrade to Pro and gain unlimited accomplishments on your resume
Instructor
Course access is free for a limited time during the DeepLearning.AI learning platform beta!
Want to learn more about Generative AI?
Keep learning with updates on curated AI news, courses, and events, as well as Andrew’s thoughts from DeepLearning.AI!