Pixeltable: AI Data Infrastructure Providing a Declarative, Incremental Approach for Multimodal Apps

To our 165+ registered workshop participants - we are so sorry to postpone our "Building multimodal AI workflows with Pixeltable" workshop with @hugobowne and @wesmckinn that was scheduled for this week. We have set a new date at the end of January, which gives us PLENTY of

Excited to collaborate with @TigrisData to make your media outputs from @pixeltablehq first-class citizens in your high-performance blob storage buckets: 🧲retrievable 🔎searchable 🩻trackable In 7 lines of #Python code

Search & retrieval shouldn’t be a maze of data plumbing. @pixeltablehq makes multimodal data feel like Postgres: declarative tables with raw data, embeddings, captions & model outputs, all queryable with SQL-like syntax. Pair it with @tigrisdata and all your data is hot. No

Github: https://t.co/f9UU1gja5i Workshop:

@a16z @JenniferHli @sarahdingwang @julesyoo @Tocelot @aleximm @venturetwins @JoshLu @stuffyokodraws @JasonSCui @MaikaThoughts This was a good read 🍿 I'm excited for 2026 to be the year companies can take unstructured and semi-structured sludge and make it just as actionable as data in spreadsheets. Full disclosure: I lead product for @pixeltablehq, a #Python open source library for multimodal

Building Multimodal AI Workflows with @pixeltablehq Dec 17 Tuesday, December 16 3:00 PM - 4:30 PM PST Marcel Kornacker (parquet, impala), @wesmckinn (arrow, pandas, ibis...), @apreshill, and @hugobowne

multimodal video processing agent :)

@kapnap_n @RayaneRachid_ Just the only Python library providing declarative data infrastructure for building multimodal AI applications, enabling incremental storage, transformation, indexing, retrieval, and orchestration of data. Not really a blackbox. It "just" unifies orchestration, storage, and

The era of text-only AI is ending. The next frontier is multimodal—video, audio, images, and sensor data that capture the world in motion. See how Pixeltable + Backblaze B2 simplify the “data plumbing” behind multimodal AI with a single interface. 🔥➡️ https://t.co/nOC2JQE50e

إذا سمعتوا عن Prompt Engineering قبل، ترى الموضوع الآن تطور وصار فيه مفهوم أهم: Context Engineering. طيب إيش يعني؟ 🤔 ببساطة: بدل ما نفكر بس "وش نكتب في البرومبت"، لازم نفكر "وش الأشياء اللي نحطها في الـ context حق الـ AI عشان يعطينا أفضل نتيجة؟". 🧵👇

@hugo1256 @akshay_pachaar You are spot on :) https://t.co/HY3IfZhcXk (disclaimer: I'm on the @pixeltablehq team)

@akshay_pachaar Treating embeddings as 'computed columns' rather than a separate microservice solves the biggest hidden issue in RAG: data drift between your source of truth and your index. Context engineering is effectively just becoming modern data engineering .

Context engineering pipeline code: https://t.co/SM2t6UkY1J Also check @PixeltableHQ's GitHub repo: (don't forget to star 🌟) https://t.co/cpWZQrltrh

It's the most wonderful time of the year to build 🎄🌟 I'm enjoying building today with @reve and @pixeltablehq (Letting my husband play prompt designer for the day)

@samhogan Would love to host the datasets with @pixeltablehq and showcase some neat data augmentation and exploration. Lmk!

Today, we're launching data sharing in @pixeltablehq Cloud: publish and replicate multimodal AI datasets with all your raw data, computed results, and embedding indexes together.

Join us next month when our guest speaker, Marcel Kornacker, Founder of @pixeltablehq, talks about data sharing on their open source platform! https://t.co/CcdXz0U6E9

@pjlbrunelle Huge respect for what the @pixeltablehq team has built. The depth of experience here is incredible. Can connect you with a few Full Stack Developers who’ve built similar systems end-to-end. Dropped a note in your DMs :)

Cut Your Data Plumbing Without Sacrificing Multimodal Power Here How 🚀 @pixeltablehq is an open source Python library that unifies declarative data infrastructure for multimodal AI workloads. thread 🧵

The @pxt.query decorator by @pixeltablehq is a powerful tool for: - AI Agent Memory: Reusable context retrieval across conversations - Multimodal RAG & Tools Systems: Sophisticated hybrid retrieval with consistent logic - Multi-Table Analytics: Complex quries as simple functions