PostgresML

Introducing the OpenAI Switch Kit🦉 🔄 Swap OpenAI with any open-source model minutes 🔄 It's an open-source AI SDK (Python & JS) that provides a drop-in replacement for OpenAI’s chat completion endpoint. ➡️ Replace OpenAI for almost any open-source model on Hugging Face in just a few lines of code ➡️ Use uncensored, highly performant, safe and customizable models ➡️ Keep AI truly open We're on a mission to accelerate #opensource excellence. Let us know how we can get you to make the switch and join us. #KeepAIopen #OpenAISwitchKit #llama #mistral https://shorturl.at/dqC68

We're excited to share how Sudowrite, the leading AI writing assistant for fiction authors, built a winning AI stack with PostgresML. Here's a taste of what they pulled off: ⚡ Functional prototype created in just 3 hours 📈 Now crunching over 1m+ calls per hour 🐘 Doing all this fancy AI stuff inside their database, with infra they already know and love Their founder's take? "Being able to co-locate your data – to query across the same metadata stack – is a no-brainer." Interested in learning how this approach could benefit your AI platform? Read the full case study here 👉 [https://lnkd.in/gqvsUis4]

Want to do RAG on websites all with one DB call? The team at PostgresML posted an amazing blog on how to use Firecrawl with their Korvus SDK which unifies the entire RAG pipeline in a single database query. Check it out: https://lnkd.in/e2ymHeFS

Too many enterprise AI/ML apps are out of date before they launch, and unmaintainable long term. 🙃 Why? Complex data pipelines. X-functional teams work across multiple quarters to get it right. Isn't there a better way? Read on in the blog. ⬇️ http://dlvr.it/TBd1Q4

RAG Retrieval Speed Showdown 🏁 We conducted a hands-on speed comparison of the most popular RAG retrieval systems: 1️⃣ Pinecone + HuggingFace 2️⃣ Qdrant + HuggingFace 3️⃣ Weaviate + HuggingFace 4️⃣ Zilliz + HuggingFace 5️⃣ PostgresML via Korvus The results? A performance gap of nearly 10x between the fastest and slowest system. 👉 Dive into our full blog post to get the details and analysis. http://dlvr.it/TBJyNN

Open source AI is not the future. It’s here, now. Hacker News has spent the last 24 hours debating if Meta’s Llama models are really “open source” rather than talking about the ramifications of its launch. They similarly debate what “AI” is. Open source AI is important, not because of some pedantic definition by some pseudo-official body like OSI, it’s important because of the power and incentive structures that pervade our society. https://lnkd.in/g5a8WH2r

"Open source AI is important, not because of some pedantic definition by some pseudo-official body like OSI — it’s important because of the power and incentive structures that pervade our society." Read on in our blog ➡️ http://dlvr.it/TB2B9p

Excited to announce meta-llama-3.1 support in PostgresML Cloud! 🦉 + 🦙 Get the details on our blog ➡️ http://dlvr.it/T9zdG2

Re-ranking retrieved results can be a powerful boost to the relevance of your response. With a managed index like PostgresML that's as easy as passing an extra parameter or two. Check out their guest post on our blog: https://lnkd.in/gcqmfzKb

RAG in a single query. On Postgres. That's Korvus. Our new open-source RAG pipeline combines the entire RAG workflow - from embedding to text generation - into a single DB query...one query to rule them all ☝️ No SQL necessary. Python, JS, or Rust. Get all the details on Korvus in our latest blog. ➡️ http://dlvr.it/T9QlcQ