Ploomber

Our editor now integrates with Google sheets. Upload invoices and push each item to a Google Sheet!

Let's build an image classifier, no coding required!

Let's build an app to help us screen resumes with AI!

Our AI editor just got better! Watch us create an app that exports bank statements to Excel in minutes!

Create AI apps in minutes with our editor! In this video, we build an app that parses bank transactions from a PDF.

I made my own Siri in 2 days, and it's better than the original.

Psst! 🤫 Missed the @Streamlit 1.45.0 release? Big news: Say hello to st.user! Learn when & how to use this new field. 👇

Ploomber MCP is available here: buff.ly/6BL2yrk

I deployed an app to the web in one sentence

Ploomber's Meerkat is now alive! The new state-of-the-art Veo 3 model is such a big leap forward! From a simple image and a prompt, we can now generate a full video with a consistent character.

PSA: Vibe Coding is dead. Asynchronous agents are all the rage now.

Cursor v1.0 just dropped 🎉 with the new Asynchronous Agent mode! Seriously impressive to build Streamlit & Dash app features with a single prompt! Check out the demo 👇

Project source: buff.ly/34mA9Ef

Is this the future of the internet? Every page is going to be generated on the fly by a LLM, based on the user's interaction with the browser.

Scammers and marketers have been overwhelming me, so I used AI to fight back. I enlisted Gemini to watch my emails, detect any spam, and waste their time. Check it out: buff.ly/nng4GWy

Building Streamlit Apps Has Never Been This Easy! Watch how you can create and iterate on apps just by chatting with Claude 4 in our AI Editor

I got tired of ordering food myself, so now AI can do it for me. Plug this DoorDash MCP server into your LLM of choice so it can make calls directly to the DoorDash API. Check it out here: buff.ly/SeT5bEs

@Panel_org We’re working on it!

@ploomber Any insights for @Panel_org users?

🔥 RESULTS: Claude Sonnet 4 is the best for Shiny R development! Tested 3 AI models for complex R apps: → Claude 4: 750 lines, production-ready ✅ → Claude 3.5: 200 lines, crashes on null data ❌ → GPT-4: Failed fetch requests, unreported errors ❌ Thread breakdown 🧵👇

🐛 Where Claude models failed: Both Claude 4 & 3.5 had identical issues: ❌ "Recovered" statistic not populated ❌ Graphs not responsive to date picker changes GPT-4: ✅ All functionality working perfectly 4/4

🔍 What we tested: Complex Dash dashboard with COVID data visualization, date filtering, and statistical displays. Key requirements: Responsive graphs, complete data filling, intuitive date controls. 3/4

New Model Review: Testing Claude Sonnet 4 vs Claude 3.5 vs GPT-4 on Dash Python! 📊 Plot twist: GPT-4 actually wins this round by implementing all functionality, while Claude models struggled. Full breakdown in thread 🧵👇