Lantern

Open-source PostgreSQL vector database to build AI applications. YC W24. GitHub:

[di qi @diqitally]

lantern team + special guest lilith :)

[di qi @diqitally]

🚀 Lantern now supports LLM calls in Postgres!

Run adhoc queries:

SELECT llm_completion('Summarize code', code) FROM files LIMIT 1;

Or automate column generation:

SELECT add_completion_job('files', 'code', 'summary', 'Summarize code');

https://t.co/RpUexnE4zU

[Umur Cubukcu @umurc]

I really enjoyed giving this talk with @diqitally at @pgconfeu. Di's blog post sums it up well, with also the code available -- a code chatbot in just 15 minutes, built with PostgreSQL.
https://t.co/iN1g0Lf2Kg

[di qi @diqitally]

Yesterday @umurc and I gave a talk at @pgconfeu about how to build a code chatbot with Postgres in 15 minutes! Here's a short version in a blog post ☺️

https://t.co/RQMq2W03WV

[Ubicloud @UbicloudHQ]

We're excited to announce managed @lanterndb on Ubicloud!

https://t.co/tjQDsYo6OU

With this partnership, here's what you get:

* With open source, powerful RAG (Retrieval Augmented Generation) AI apps
* Better better price / performance than proprietary solutions
* Usability and fast indexing features from Lantern
* Deep Postgres and AI expertise, in one partnership

[Umur Cubukcu @umurc]

Building AI apps with Postgres is now a lot simpler -- and the apps are that much faster! 🚀

It is wonderful to welcome @lanterndb as a partner on Ubicloud, where our users get 1st party support directly from the creators of modern, powerful @PostgreSQL extensions. More thoughts on this on our blog!

[di qi @diqitally]

So excited about this partnership! 🥳

@UbicloudHQ is building an open source alternative to AWS that reduces your Cloud bill by 3-10x

It’s been wonderful to work with the Ubicloud team. They’re experts in Postgres — they're the same team that built distributed Postgres at Citus

[Lantern @lanterndb]

🚀 We’re excited to share that Lantern is now available on @UbicloudHQ !

You can access vector generation, LLMs, serverless indexing, and more. All in Postgres. All open-source.

If you're building an AI app, we can help you simplify your infrastructure.

https://t.co/oomWoZ2f1R

[andre --dangerously-skip-permissions @andrezfu]

Vector Databases & Document Intelligence in the same blog post? What's that - a cross over episode?!

Super excited to announce our cross-blog with @diqitally from @lanterndb along with the greatest CTO on Earth @nankigrewal.

In this post we take a look at how to automate Banking Compliance with Ecliptor's document intelligence and Latern's vector database capabilities.

Don't want to spoil your fun reading it, so check it out below 👇

[di qi @diqitally]

we were excited to work with @nankigrewal and @andrezfu from our yc batch on this demo of automating compliance checks with lantern and ecliptor! 🚀

we used ecliptor for document processing and lantern for vector generation + storage. check it out ☺️

https://t.co/nASVvCP0gG

[di qi @diqitally]

we went to an @OpenAI x @ycombinator hackathon this week!

@ngalstyan4 and i built an app to iteratively build postgres extensions - it generated a median aggregate function extension successfully :)

[Daniel Chalef @danielchalef]

This is really cool. A super clever way of reducing pgvector’s indexing impact on production database use.

[di qi @diqitally]

We built external indexing for pgvector in Postgres.

External indexing offloads the resource-intensive index creation process to external machines to reduce the impact on database performance.

Read the blog post from @D4RK7ET below 👇

https://t.co/1r8J2Nyudg

[di qi @diqitally]

I built a site to browse the Postgres mailing list archives. It supports search using FTS from Postgres + vector search with the Lantern extension.

I've been using it for testing but cleaned it up a bit to share 🙂

https://t.co/b0NcjqwWp7

[Sam Willis @samwillis]

This is a very cool look at how pgvector indexes work, and they built an incredible browser based visual inspector for it with PGlite 🤯

[Varik @D4RK7ET]

Besides the deep dive we have built a JSON parser for HNSW indexes and interactively visualized index pages entirely in the browser using pglite.

Check out the blogpost and play with the index parser yourself!

#PostgreSQL #postgres #pgvector #hnsw #pglite

[Lantern @lanterndb]

We did a deep dive into how pgvector stores HNSW indexes in Postgres.

Read more from Varik below ↓
https://t.co/bbrc5cuQts

[Lantern @lanterndb]

Did you know there are over 300 parameters used to configure a Postgres service? @ngalstyan4 covers some of the differences in default settings across Cloud providers.

https://t.co/8CPq1ka4DD

[Lantern @lanterndb]

Hybrid vector search = dense vectors + sparse vectors, combining semantic + keyword search to improve search results 🚀

We evaluated hybrid vector search with embeddings from Open AI, @JinaAI_ , and SPLADE using the BEIR framework.

https://t.co/L8IHcWUKCU

[Lantern @lanterndb]

We respond to Pinecone's recent blog post comparing Postgres and Pinecone. We show that Postgres can outperform Pinecone in the same benchmarks Pinecone covered in their article.

https://t.co/UmP0fKye8F