We're excited to share this recent podcast by Ben Lorica (Ben Lorica 罗瑞卡) where deepset co-founder Malte Pietsch (@malte_pietsch) talks in depth about the power of our Haystack framework. 🚀 🚀

thedataexchange.media/orchestration-…

#LLM s Haystack #LLM

🚀 Building apps with large language models can be tricky. Malte Pietsch deepset on how their open source Haystack framework makes it easy to orchestrate #LLM s , data stores, and pipelines so you can focus on your app logic, not infrastructure
#LLM
thedataexchange.media/orchestration-…

Just a few days ago, OpenAI announced JSON mode. And a few hours later Haystack is ready to support format enforcement, all with open-source tooling. Welcome to the LM Format Enforcer which just joined our Integrations page 🫶 and thank you to Noam Gat for this great…

The retrieval step for RAG pipelines is one of the most important steps to get right. But what happens when dealing with multilingual data such as hotel reviews?

The answer is multilingual embedding models 🌎

Check out the new article by Bilge to learn how to build a…

👀 In case you missed it live, our workshop with Haystack is now available for replay. RAG, agents, and memory come together to create high-performance Conversational AI – tune in and learn more: hubs.ly/Q0283QRD0

Haystack 1.22 is fresh out the oven 🧑‍🍳

• New features to preview Haystack 2.0 in the preview package!: ChatMessage, HTMLToDocument, HuggingFaceLocalGenerator and more!
• All Pinecone index types are enabled, including Starter
👇
github.com/deepset-ai/hay…

It's great seeing newer and better open-source LLMs and testing them out with Haystack. The latest Zephyr models by Hugging Face were tested out by Tuana and Stefano Fiorucci using the preview for Haystack 2.0. You can read their full guide here 👇
haystack.deepset.ai/blog/guide-to-…

I'm hosting a Haystack workshop at the Anthropic Hackathon at 4:30 PM 👇

1️⃣ Interacting with Claude using Haystack
2️⃣ Building Retrieval-Augmented Generative Pipelines
3️⃣ Creating Custom Components with Haystack for custom RAG applications
4️⃣ Indexing Pipelines with…

While generative models like GPT models are invaluable for a wide range of natural language tasks and trained on extensive text data they do come with a significant limitation. They have a knowledge cutoff point, which marks the end of their training data and restricts them from…

How does RAG handle situations when the information it retrieves doesn't have the answer? Here are some examples:

1. Default Response for Empty or Insufficient Retrieval: If the Retriever returns an empty result or the retrieved information is insufficient, the LLM may not have…

Check out the ⚡️lightning talk I gave at the The AI Conference last month:
Taking retrieval-augmented generative pipelines with Haystack to the next level with document ranking 🫶

A couple of things about the video:
1. Say hi to Philip (make love not war🌈) 😅 I love the little head bop up at…

🎃 Hacktoberfest 2023 has been a phenomenal journey for us at Haystack! 🙌 Huge thanks to all our incredible contributors who joined us to celebrate the open source and poured their talent into Haystack. We're already looking forward to next year and the amazing things we'll…

The Haystack PromptNode is super flexible. It allows you to connect to and prompt a number of LLMs. It accepts flexible prompt templates which define the manner of interaction with LLMs. And we're really excited for you to make use of this component in the upcoming Anthropic…

AI's role in publishing? Empower, not replace. Learn how LLMs are transforming the industry!

deepset.ai/blog/artificia…

#AIinPublishing #LLMs

Join the deepset, Cerebral Valley, LangChain and Balderton Capital teams for a day of LLM hacking in London 🇬🇧 4-5 Nov

The Anthropic Hackathon is an opportunity to get your hands dirty with some code and build LLM applications with Claude and Haystack
partiful.com/e/pQHQrWPg1A6P…

Searching multiple #OSS project docs at once? Check out the new demo by Silvano Cerza with Haystack 2.0: It combines DocArray, streamlit, #Jinja , #Pandas , Elastic, and #NumPy docs to generate answers with references.

Ready to give it a whirl? 👉
huggingface.co/spaces/deepset…

📆 Join us on November 2, at 5 PM CET/8 AM PDT for a deep dive into Conversational AI 🗣️💬

In this webinar, we will delve into RAG, agents, and memory - how to integrate them all using Pinecone and Haystack to create conversational AI 👇
pinecone-io.zoom.us/webinar/regist…

Feature appreciation post. Sometimes, code can get complex. So why not simply draw? ✍️🎨
With Haystack 2.0, pipelines will be very flexible. Sometimes, it's nice to quite literally look at them!
An example of a RAG pipeline with a recentness ranker👇
Simply: pipeline.draw()