🚀 BIG NEWS: We just shipped @platformatic/python to help you run Python ASGI apps INSIDE your Node.js process! This changes everything for AI/ML + Node.js apps 🧵

❓ "Wait, Python... inside Node.js? Why?" Here's the thing: Every time your Node.js frontend calls your Python ML service, you're paying the network tax. What if that call took microseconds instead of milliseconds? What if it couldn't fail due to network issues?

🏗️ How it works - Embedded Python interpreter runs in your Node process - Rust bridge handles the translation (zero network overhead!) - Direct memory sharing between JavaScript and Python threads - Your FastAPI/Django app runs unchanged No sockets. No serialization. Just speed

⚡ Real benchmarks: - 5,200 req/sec with @platformatic/python - Beats fastapi run, daphne, and hypercorn - Sub-2ms latency - Zero network overhead Your ML inference just got 10x faster. Your data pipeline? Same story.

🎯 Use cases that are now possible: - Real-time fraud detection with PyTorch models - AI chatbots with truly instant responses - Image processing without API calls - Pandas data analysis in your Next.js app - LangChain + Node.js in one process

💻 Getting started is simple. Drop your Python app in your public docroot and then:

🔥 Or use Watt for the full experience: npx wattpm@latest create --module=@platformatic/python This gives you routing, config management, hot reload, and everything else Platformatic offers - now with Python superpowers.

🧪 We tested a Next.js + FastAPI combo app: Frontend: Next.js Backend: FastAPI Both in ONE process. Deployment? One container. Monitoring? One service. Scaling? Together. Check it out:

🛠️ The secret sauce: - Rust-powered http-handler crate for zero-copy bridging - WebAssembly-based binary patcher for Python library loading - Thread pool management for Python workers - Full ASGI protocol support This isn't a hack. It's production-ready engineering.

🚀 What's next: - WebSocket support coming - Deeper streaming capabilities - Enhanced observability between runtimes - Even more performance optimizations This is just v1. The future is polyglot applications running at native speed.

🎉 @platformatic/python is live NOW on npm! What will you create? 🔮 Read the full announcement:

Go build something incredible. Mix PyTorch with Next.js. Run Django inside Fastify. The barriers are gone. If you have any questions, join our community on discord:

Wow this looks really cool and much better than the hacky thing I was planning to work on 😂

You totally spot what we were building! Thanks for validating the idea!

This is super hot! Nice work Platformatic team. Way to bring the best services to all developers! 👏

Hey, your reactive programming intro reshaped big fan! I built Nexus Backpressure, a protocol to unify backpressure across RxJS, Kafka, Node Streams, etc., with ~100 LOC core and adapters. Curious if it resonates. Check it: Thoughts?Daouda Abdoul Anzize

check dm

boom

Can this be used as runtime for any python code? Ala IPython kernel for Jupyter?

nice timing! I just need to use some models from Python