Datalayer

6/6 Check out our official docs here ➡️ core.datalayer.tech/python/

5/6🧹 Everything’s wrapped in Python context managers, so resources auto-clean when you’re done.

4/6 📸 Capture full session snapshots for reproducibility. Pause, share, or pick up right where you left off.

3/6 ⚙️ Create and manage remote CPU/GPU runtimes directly from your notebook or Python script. This allows you to scale your workflow on the go within a Python interface.

2/6 🔐 Instant authentication, just set one env var and you’re in. So you don’t have to struggle around with tokens or crazy configs.

1/6 🐍 Datalayer’s Python SDK is here to bring cloud compute management into your data and AI workflow

Here’s how it works 🧵⤵️

Tried GitHub Copilot in VS Code? How about an AI Agent for Jupyter Notebooks? I’ve spent more time coding and digging into stackoverflow for doing python for my notebooks than actually working on my research and…

7/7 Read our latest article to learn more about Jupyter AI Agents:

🔗 jupyter-ai-agents.datalayer.blog/tried-github...

Datalayer Ξ AI Platform for Data Analysis Datalayer Ξ AI Platform for Data Analysis.

6/7 Get started today:

1️⃣ Sign up at datalayer.ai (and receive 50 free credits)
2️⃣ Create a notebook & assign a remote runtime to it
3️⃣ Type your prompt & watch the AI do the its job 🤖

5/7 This way, you can focus on insights and making progress in your project while AI handles the coding and writing the notebooks.

4/7 Everything runs on a remote GPU/CPU without having to change your dev environment.

3/7 Now using AI Agent on Jupyter you can just type your request on natural language and let the AI agent:

❇️ Writes code & markdown
❇️ Embeds charts, tables & notes
❇️ Auto-cleans failed cells

Datalayer Ξ AI Platform for Data Analysis Datalayer Ξ AI Platform for Data Analysis.

2/7 You might be spending hours debugging Python in notebooks and having StackOverflow open in multiple tabs. 😩

✨ Meet Jupyter AI Agents (1/7)

🧵⤵️

4/ Official docs here: docs.datalayer.app/secrets/

3/ Manage secrets from the UI (under the settings tab) or with a datalayer secrets command in the CLI.

2/ Store your API keys, psw, and other creds in our vault. They auto-inject as env vars in every runtime.

1/ v1.1.1 of Datalayer Core is out, and it got Secrets Management 🔐

🧵⤵️

GitHub - datalayer/jupyter-mcp-server: 🪐 ✨ Model Context Protocol (MCP) Server for Jupyter. 🪐 ✨ Model Context Protocol (MCP) Server for Jupyter. - datalayer/jupyter-mcp-server

3/ If you haven’t starred us yet, feel free to check out our github repo below and give us a ⭐ Any support helps others find the project.

🔗 github.com/datalayer/ju...

2/ In case you missed it, Jupyter MCP Server is like having an AI sidekick in your notebook. Imagine real-time edits, auto-retries on failures, and smooth (as 🧈) integration with Claude Desktop, Cursor, Windsurf, and pretty much any other MCP client you’re using right now.

1/ 🚀 We just crossed 500+ stars on GitHub for our Jupyter MCP Server repo!

Big thanks to everyone who’s starred, contributed, or shared their feedback.

🧵⤵️

Base Python SDK is now available - Read the documentation on core.datalayer.tech/python/

You can authenticate, create runtimes, execute code, take snapshots with a few line of Python code 🐍

The Big LLM Architecture Comparison From DeepSeek-V3 to Kimi K2: A Look At Modern LLM Architecture Design

The Big LLM Architecture Comparison, by the excellent @sebastianraschka.com

magazine.sebastianraschka.com/p/the-big-ll...

datalayer/jupyter-ui ⚛️ React.js components 💯% compatible with 🪐 Jupyter https://jupyter-ui-storybook.datalayer.tech - datalayer/jupyter-ui

today achivement: pruned the jupyter-ui open issues, going from 89 to 42 github.com/datalayer/ju...

A lot of them related to next.js which is now way better supported 🚀

YouTube video by Cloud-Native Geospatial Forum [CNG 2025] AI Agent & MCP Servers for Jupyter to simplify Earthdata Discovery/Analysis –Eric Charles

🆕 #CNG2025 session recording
@echarles.net from @datalayer.ai shows how to use their AI platform to write and run Jupyter notebooks for you! This means less time coding, more time on insights.
▶️ youtu.be/cu1pOCeR-9Y

5️⃣ Wanna try it out? Check the docs over here:
jupyter-mcp-server.datalayer.tech/datalayer

4️⃣ So everybody in your team can see the same live view.

3️⃣ No more changing tabs or syncing files. One interface, one notebook.

2️⃣ Outputs, charts, tables, everything streams back into your editor without context switching or juggling tabs.

1️⃣ Imagine writing code in VS Code (or Claude Desktop, Cursor, etc.) and your cloud notebook updates in real time.

📢 UPDATE: Jupyter MCP Server now supports Datalayer Provider.

🧵⤵️