Building Your Intelligent Assistant with Hugging Face and Elasticsearch

on YouTube ↗02h 51m

" summary

Shows a concrete pipeline that crawls and indexes about 2,900 Devoxx talks into Elasticsearch, computes 1536‑dim dense embeddings plus sparse embeddings, and performs hybrid KNN + keyword retrieval. The retrieval results feed Hugging Face / OpenAI models for retrieval‑augmented generation (RAG) summarization to reduce hallucinations by citing sources; tested models include BERT, Falcon/Mistral families and Whisper for speech, and a voice prototype with a wake‑word demo.

AI-written summary · about

" chapters (17)

1. 0:00Introduction
2. 1:04Speakers and context
3. 3:13Agenda
4. 12:46NLP and assistants history
5. 27:36Search engines to generative AI
6. 37:50ChatGPT as assistant
7. 46:07Question Answering and RAG
8. 57:23Voice processing
9. 1:00:57NLP and search engines
10. 1:14:11Vector search and embeddings
11. 1:27:58Hugging Face overview
12. 1:41:18Hugging Face demo
13. 1:57:22Elasticsearch and Hugging Face
14. 2:01:45Elasticsearch and NLP demo
15. 2:30:21Hybrid search and vectorization
16. 2:41:26Final RAG demo and summary
17. 2:47:17Resources and conclusion

" related talks

• 33mMachine Learning 2.0 with Hugging FaceJulien Simon · Conf42 Machine Learning 2022
• 01h 48mInformation Retrieval from the Ground UpPhilipp Krenn · AI Engineer World's Fair 2025
• 49mMachine Learning 2.0 With Hugging FaceJulien Simon · PyData London 2022
• 44mDecoding generative AIAlex Coqueiro, JD Lynch · AWS Public Sector Symposium Ottawa 2023
• 28mUnlock the Power of LLM: Build a HuggingFace AgentDarya Petrashka · Conf42 Python 2024
• 19mBuilding Machine Learning Applications using Hugging FaceLewis Tunstall · Uphill Conf 2024
• 24mEmpowering Vue Apps with AIAileen Villanueva Lecuona · VueConf US 2024
• 21mBuilding RAG With OpenSearch ML Plugins: From PDFs To Voice-Enabled SearchKushagra Sharma · OpenSearchCon Europe 2026