How to build RAG pipelines fast?
Graphlit founder Kirk Marple babysits Paul on understandig how to leverage the best RAG platform.
👣 Podcasts never show code, demo videos are made by people who are already experts in their field:
The human Co-Pilot series merges both, providing a format where the beginner (Paul) gets to ask the expert (Kirk) any silly questions.
This should help the understanding of developers new to the technology, by following the journey of another noob.
📝 In this teaching sessions, Paul learns about ingestion, indexing, data preparation, extraction, enrichment, specifications, and more.
In the second half of the video, Kirk also explains the concept of GraphRAG, and how easy it is to start building Graphs using RAG-as-a-service platform Graphlit.
🧠 Paul really sees the potential with Graphlit because it eliminates typical RAG setups:
single API Key for Graphlit instead of 2 or more
No need to write embedding code using Cohere or OpenAI
No need to write upserting code using Pinecone or Weaviate
Default chunking, Embedding and storage options of Graphlit are great for beginners, but option to make it much more sophisticated
🇪🇺 If you're reading this, wanna use Graphlit and you're from Europe, please reach out to Kirk to ask him to set up a Graphlit instance in the EU.
Right now there's not enough demand for it, but if enough people express interest, we'll get the EU instance!
📌 Timestamps:
0:00 - What is graphlit
2:30 - Set up a graphlit project
6:12 - Ingest first piece of content
12:10 - Graphlit stages expalined
15:30 - Translate the docs to code
21:42 - Paul slowly gets it 🎉
27:42 - Graphlit processes explained
30:32 - Difference RAG and GraphRAG
32:03 - Run what we have so far
36:48 - How to access ingested content
39:54 - Steps needed to enable GraphRAG
45:00 - Specification walkthrough
47:27 - Conclusion (Graphlit is awesome)
🔗 Links and Resources:
Graphlit: graphlit.com
Notebook from the video: https://colab.research.google.com/git...
Follow Kirk on X: https://x.com/KirkMarple
DentroAI - AI Integration and Implementation for SMEs: https://dentroai.com
Follow Paul on X: https://x.com/paul_dentro
DentroChat (multi-LLM chat): https://dentro.chat
NoteThisDown (handwritten Notes to Notion): https://note-this-down.com
#livecode #pythontutorial #rag #codingtutorial #developershowcase #ragchatbot #graphrag #liverecording #techinterview #fromdevstodevs #dataextraction #jupyternotebook #founders #graphlit #ragasaservice #tech #python #backend #api #airag #llm #retrievalaugmentedgeneration