📆 Get help to implement this workflow: https://cal.com/vladimir-de-ziegler-uyu7qb/15min
🤖 Source code: https://winning-crafter-306.kit.com/24c2e2dca0
🐡 Thanks to AquaVoice, the sponsor of this video: https://withaqua.com/launch
🛠️ What you will learn in this video:
This is a full tutorial on how to develop AI agents to handle your claims. I’m really excited about this as I’ve been getting a lot of requests to build something like this using CrewAI.
By the end of this tutorial, you’ll know how to:
✅ Extract info from complex PDF documents using OCR technology (docling)
✅ Make your long form documents searchable using LLMs (llamaindex)
✅ Build RAG pipelines to query your documents (llama index)
✅ Build custom tools for your agents to interact with your data
✅ Build a team of agents to analyse documents and come up with insights and recommendations
✅ Set up a workflow that integrates with your existing team to make them more efficient
This workflow can be applied beyond the insurance industry. Whenever you have a set of documents you need to compare to guidelines, regulations, etc. it can work. Be it KYC to onboard new customers, checking sample results against criteria, compliance, etc.
📰 Stay updated with my latest projects:
📍 LinkedIn: /vladeziegler
🐦 Twitter/X: / ziegler_de
✉️ Email:
[email protected]
Who am I ?
Hi, I’m Vlad. I build AI agents for entrepreneurs who want to 10x their output without growing their teams.
Book your free audit now: https://cal.com/vladimir-de-ziegler-uyu7qb/15min
⏰ Video Timestamps:
0:00 Intro
0:50 Agenda
2:00 Breakdown of the workflow (claim management agent)
3:16 Part 1: Making your data LLM-ready (parsing, vectorisation)
9:00 Part 2: Building custom tools for your agents (RAG pipelines)
14:57 Part 3: Building the AI agent (CrewAI)
20:12 Demo
🎥
If you want to dive deeper into certain topics, I have additional tutorials:
- CrewAI deep dive:
https://youtu.be/g39NxLNNu1M?si=8BnbmGcE5C7i5U1k
- Vector embeddings deep dive:
https://youtu.be/ggo781QGYVc?si=M8Wj7pfgMzkdt2-9
Or other workflows:
- Build your Data Analyst Agent:
https://youtu.be/FiAYC2_NqIo?si=_M0WO7gGpHmjsFP7
- Build your Sales Agent:
https://youtu.be/ggo781QGYVc?si=M8Wj7pfgMzkdt2-9