Weaviate Podcast

Hey everyone! Thank you so much for watching the 37th episode of the Weaviate podcast! This episode discusses some of the ideas behind GPT Index. GPT Index presents really exciting ideas about how we use LLMs to index our data and then traverse these data structures. We began the podcast by discussing the origins of the tool and the ideas behind the Tree Index. We then discussed generalizing these trees to graphs and whether we are headed to the Knowledge Graph 2.0. Another really interesting topic we covered is the inference cost of building and traversing LLM indices like this! I really hope you enjoy this podcast I think these are some of the most cutting edge ideas in AI and Search! Check out GPT Index (now LlamaIndex here - https://gpt-index.readthedocs.io/en/l...) Chapters 0:00 Introduction 0:18 Origin Story of GPT Index 2:22 GPT Tree Index 5:53 Search Examples - Podcast Clips 11:22 Knowledge Graph 2.0? 16:05 LLM Writing Data to DB 20:18 Weaviate Classes and Index Hierarchy 23:53 Subindices vs. Tool Use 28:50 Inference Requirements for GPT Index 35:53 Design of GPT Index 37:40 Impact of Cheaper LLMs for this 40:02 Name Change for GPT Index? 42:04 Llama Hub 45:07 Relationship in Software Stack 48:15 Extension to Multimodal, e.g. Vision-Language

Hey everyone! Thank you so much for watching the 36th episode of the Weaviate podcast! This episode continues on the marriage between LLMs and Semantic Search, welcoming back Weaviate CEO and Co-Founder Bob van Luijt! Enter LangChain and its creator, Harrison Chase, providing the glue between LLMs and tools, such as semantic search. LangChain provides a set of abstractions around chaining multiple language model calls with different prompts, strategies for overcoming the 4096 token limit, and connecting LLMs with their tools. LangChain Hub is a collection of these chains if you want to check it out yourself! Huge thank you to Harrison and Bob for joining the podcast, this was such an information packed podcast with some great predictions for the future of LLMs + Vector Databases!   Check out LangChain here! https://langchain.readthedocs.io/en/latest/ Chapters 0:00 Welcome 0:14 Origin Story of LangChain 1:27 What are LLM Chains? 4:00 Adding Weaviate Search 7:30 LLM Orchestration and Tool Use 11:24 Extension to Multi-Modal 14:00 Natural Language Interaction with Software 20:36 Will Prompt Engineering Last? 21:00 More on Tool Use 25:47 Favorite Prompts 29:54 Temperature in LLMs 31:00 Reasoning and Knowledge 32:50 LLM as Router 35:50 Model Diversity 39:45 No GPUs before PMF 41:35 Virality of LangChain 43:40 Future of LangChain

This podcast debuts a huge new release from Weaviate... the generate module! The generate module is a new API in Weaviate that facilitates passing YOUR data from the Weaviate database to ChatGPT. This enables ChatGPT to become knowledgeable about your particular business or interests! Here is a great snippet from Bob around the 43 minute mark that describes how this kind of LLM technology is changing the world of database technology, "Yeah so, what I’m really excited about and this is something that it’s just so funny right because if you see it, you have this huge epiphany. I’ve always been thinking of working with these models on input. Right so that they we can solve the problem of not having 100% keyword based search, so that we can have semantic search, image search, and those kind of things. I saw that as this beautiful uniqueness coming from a vector search engine or vector search database. So now what we’re adding is not only the input in the database but the output. So we’re basically saying we’re going to give you relevant information coming from the database, but that’s not per se stored inside the database. That’s new! I mean, just think about the most used databases in the world, Postgres, or MySQL, those kind of databases. It only outputs what’s in there. It makes sense. Because that’s how you use it. But now what we’re saying, is that’s fine you can do that, but also it can give you information, give you data that’s generated based on a task or prompt that you’re giving it. Having databases that make sense of it at input and generate new relevant content if that’s something you want as a user is amazing, and it’s just getting started. We should do this podcast like a half a year from now again and see how it's evolved because this is just too exciting man.". I really hope you enjoy the podcast, we are more than happy to answer any questions or help you get started with Weaviate!

Chapters 1:00 Introduction 1:26 Why does the world need yet another database? 3:57 Memory Allocations 9:40 Delayed Decoding 16:05 SIMD 22:04 Demo Time! 24:38 Mad at Go? 26:00 Audience Questions

I am so excited to host Dmitry Kan on the Weaviate Podcast!! Dmitry is a world class expert on emerging trends in search technology! This podcast reflects on Dmitry's latest characterization of the field, the Neural Search Pyramid. This describes the different components involved with building a Deep Learning-powered Search experience from the Approximate Nearest Neighbor index algorithms, to Database functionality,  LLM orchestration, Vectorization optimization, Data preprocessing, User Interface, and many more! We also concluded the podcast with an interesting debate around renaming "Vector Search" to something else that reaches a broader audience. I really hope you enjoy the podcast, thank you so much for listening! Please see the links below to Dmitry's recent content and the Weaviate Podcast Search App! Links: Dmitry's Keynote at Haystack Europe 2022, Where Vector Search is Taking Us - https://www.youtube.com/watch?v=2o8-dX__EgU Dmitry's latest blog post on Neural Search Frameworks: A Head-to-Head Comparison - https://dmitry-kan.medium.com/neural-search-frameworks-a-head-to-head-comparison-976aa6662d20. Search through this episode of the Weaviate Podcast! - https://github.com/weaviate/weaviate-podcast-search Chapters 0:00 Neural Search Pyramid Visual 0:40 Weaviate Podcast Search! 1:35 Welcome Dmitry!! 2:02 Where is Vector Search taking us? 5:40 Retail and Search 11:02 Neural Search Frameworks 17:10 Data Preprocessing, e.g. PDF to Text / OCR 24:15 Vectorizing Data 31:18 ANN Index and Database Entanglement 37:25 Hardware Accelerators for Vector Search 46:02 Reader Layers, Q&A, Ranking, … 51:20 ChatGPT in Neural Search Frameworks 1:03:40 Search Result Summarization with ChatGPT 1:12:55 User Interfaces for Neural Search 1:26:30 Renaming “Vector Search” 1:46:10 Thank you Dmitry!!

Weaviate podcast #33. Thank you so much for watching the 33rd Weaviate Podcast! This episode features one of the heroes of Deep Learning for Search, Nils Reimers! Nils' work on SentenceBERT is one of the foundational works for applying Deep Representation Learning to text search. This is the idea that personally inspired me to work in this field. Having seen the successes of Contrastive Representation Learning for Computer Vision, I was mind-blown by the possibility of this for NLP and text search. In addition to the scientific foundation, the software development of the Sentence Transformers library and BEIR benchmarks has been enormously impactful! It was an honor getting to ask Nils the questions I have about these things, from the role of Data Quality to Intent, Sparse Vectors, Long Document Encoding, Distribution Shift, and many more. I really hope you enjoy the podcast! We are so excited about the Cohere Multilingual embedding model and can't wait to see what else comes out of Cohere and their amazing team! Cohere Multilingual ML Models with Weaviate: https://weaviate.io/blog/2022/12/Cohe... Nils Reimers: https://scholar.google.com/citations?... Mentioned in the podcast, Cross-Encoders: https://weaviate.io/blog/2022/08/Usin... How to choose a Sentence Transformer from HuggingFace: https://weaviate.io/blog/2022/10/How-... Chapters 0:00 Cohere X Weaviate 0:22 Welcome Nils Reimers! 1:18 Origin Story 3:15 Learning Text Embeddings 6:54 Positive and Negative Sampling in Contrastive Learning 13:32 1 Billion Pairs for Text Embedding Optimization 15:44 Impact of Data Quality 18:40 New Cohere Multilingual Model! 24:50 Challenge of Debugging Multilingual Models 28:30 Intent in Search 30:40 Thoughts on ColBERT 33:50 Sparse Vectors in Search 36:17 Long Documents and Multi-Discourse 43:40 Entity Parsing in Query Understanding 46:08 Unknown Words and Distribution Shift 50:07 Re-Vectorizing with Fine-Tuning 53:07 More on Search Interfaces and Intent in Search 55:15 Thank you Nils!

Weaviate Podcast #32. Thank you so much for watching the Weaviate podcast! We are super excited to host Sam Bean from You.com! As well as welcome Zain Hasan and John Trengrove to the Weaviate podcast for the first time! Sam begins by describing You.com, and then we dive into the Weaviate Spark Connector that Sam played a massive role in creating. I thought this was such a masterclass in the Spark big data technology; John, Sam, and Zain are all data engineering pros and I've never learned more about a new technology from a podcast than this one. I really hope you enjoy listening to it, please let us know any questions or ideas you have. Also, please see Zain's blog post on "The Details Behind the Sphere Dataset in Weaviate" - https://weaviate.io/blog/2022/12/deta.... This provides great detail on exactly how to use the Spark connector in Weaviate! In this case for a billion-scale dataset upload!!! Chapters 0:00 Thanks for Watching! 0:18 Welcome Zain and John 0:28 Welcome Sam Bean, You.com! 1:48 Search Interface and Search Apps / Widgets 3:40 Searching through Specific Websites 4:00 Origin Story of You.com 6:53 How did you come across Weaviate? 8:33 Text, Image, Audio Search 10:28 What do you use Spark for? 14:20 Datasets used with Weaviate 16:14 Creating a Spark Connector to Weaviate 21:05 Adding Streaming support 22:50 Vectorizing Data at You.com 27:15 More on ONNX + Spark 29:52 Performance Questions, Spark + Weaviate 34:35 Parquet for HuggingFace Dataset Files 34:54 What is Parquet? Spark Pushdown Filters Explained 39:04 Similar to HDF5? 39:45 Spark for Extracting Ranking Features 43:25 Hybrid Search 46:48 Collecting Search Relevance Data 51:07 Thank you for watching! Thanks Sam, Zain, and John!

Weaviate Podcast #31.  Weaviate 1.17!! This is a massive release for Weaviate, debuting Replication, Hybrid Search, BM25, Faster Startup and Import Times, as well as other fixes! Replication and Hybrid Search are two massive features for Weaviate, and we really hope you enjoy the description of them from the podcast. Please also check out the Weaviate 1.17 release blog post for more information as well - https://weaviate.io/blog/2022/12/Weaviate-release-1-17.html!   This is also a very special podcast as we welcome Parker Duckworth for the first time to the podcast! Parker gave an excellent explanation of Replication and unpacked some of the questions we are seeing around Ref2Vec!  Thank you so much for listening to the podcast! Please check out the newest version of Weaviate!   Chapters  0:00 Weaviate 1.17! Welcome Parker!  0:28 From Italy to 1.17  2:04 Replication work in Italy  3:58 Replication Details  6:28 Use Cases of Replication  13:12 Product Engineering  16:24 Hybrid Search  21:30 Open Question around Hybrid Search  23:15 Rank Fusion 24:00 BEIR Benchmarks  27:28 What is Ref2Vec?  29:08 Bipartite Graph Ref2Vec Example  29:30 Graphs in Weaviate  34:25 Ref2Vec Cascading Updates  37:45 Custom Aggregation Functions in Ref2Vec  39:08 Adding Recency Bias in Ref2Vec  41:18 Startup Time Improvements  41:50 Batch Latency Improvement

Weaviate Podcast #30.  Chapters 0:00 The future of search! 0:42 Welcome Marco and Chris! 4:28 Solving Hallucination with External Memory LLMs 8:16 Bob van Luijt on Weaviate and LLMs, Collaborations 14:48 What we have is not yet what the technology is capable of 16:45 Everything is Search! 18:55 The Magic of Machine Learning 20:30 Asking follow up questions 22:28 Meaning in LLMs and RLHF 27:10 How ChatGPT is Evangelizing the Technology 29:45 What is the future of search from a user perspective? 34:38 Integration with Existing Businesses 35:20 Impact on Creativity 37:37 Data Visualization from Natural Language Questions 39:00 Thought experiment - is this notebook an extension of the brain? 42:20 More on “always on” interface 43:42 General vs. Specific Intelligence 45:25 Software Business Impact 48:12 Open-Source Models 49:25 Finding Niches 56:30 Pride in Humans + AI 57:20 Exploring more Prompts 59:20 Personalized Embedding Key 1:02:40 Concluding Thoughts

Weaviate Podcast #29. Hey everyone, thank you so much for watching another episode of the Weaviate podcast! This episode features Matthijs Douze, one of the most talented and accomplished scientists we've hosted on the Weaviate podcast! Matthijs has pioneered the use of Product Quantization to compress vector representations and enable even faster and more efficient approximate nearest neighbor vector search. Matthijs told an incredible story about the history of this research, from searching from SIFT vectors for Computer Vision Search applications like real-time CD Cover album search to the problems facing modern IVF-PQ systems and the use of PQ in graph-based HNSW search. This is also a very special episode as Abdel Rodriguez makes his debut on the Weaviate podcast to discuss Weaviate's efforts in integrating PQ support and the unique challenges with this algorithm and the incremental updates required for a Vector Database. On this topic, Etienne Dilocker also returned to discuss the topic of Vector Database vs. Library with Matthijs, who is one of the lead developers of the Faiss library. This was a really information-heavy podcast, please don't hesitate to ask us any questions or present any of your ideas! Thanks again for listening!