Profile avatar
ssp.sh
Dad. Technical Author, Data Engineer. Data practitioner (20y) • Writing at ssp.sh since 2015. Focused on the craft of data engineering & storytelling. 📚 vault.ssp.sh • 📖 dedp.online ❯ #dataengineering, #opensource, #writing, #obsidian, #neovim
3,507 posts 4,962 followers 1,089 following
Regular Contributor
Active Commenter
Posts 14 Comments 36

Today's event was in Zurich, Switzerland, at the Google Office for the Applied GenAI community. It was great to connect. For the curious, the slides for my lightning talk about «SQL Meets GenBI: AI-Powered Dashboards in a Declarative World» are on www.ssp.sh/slides/.

submitted 6 hours ago • 1 comment

What is your take, specialize, or generalize, or both? 🤯😉

submitted 14 hours ago • 2 comments

Why do you almost always need another lawyer to make your own legal expenses insurance pay, even though you pay them to work for you.. 🤷🏻‍♂️

submitted 1 day ago • 0 comments

Is anybody using WhatsApp Channels? Not using them myself, but if there is some interest, I can upload the latest there. 💬 whatsapp.com/channel/0029...

submitted 2 days ago • 0 comments

Trying Rill MCP implementation with Claude. It worked on the first try with a local Rill and NYC and Pickup data set. I think we arrived at "self-serve BI" after many years of trying if we roll out MCP capabilities to everyone in the org. Beyond mere numbers, we get interpreted context, too.

submitted 2 days ago • 2 comments

To yesterday's 📢: #DuckLake is a @duckdb.org's approach to simplifying Lakehouse architectures. It uses a standard SQL database (DuckDB) for all metadata instead of file-based systems while still storing data in Open File Formats. The goal: To make it more reliable, faster, and easier to manage.

submitted 6 days ago • 2 comments

New 📔: Data storage costs are eating into analytics budgets. Wouldn't it be great to build a data warehouse on top of affordable storage? SSDs are fast storage but expensive. S3/R2 is significantly cheaper, but the downside is that it quickly becomes messy and lacks clear governance and rules.

submitted 11 days ago • 1 comment

What's the best tool to create presentations in Markdown? Requirement: They can export to PDF and still look visually appealing. Here's the primary open-source slides creator for markdown I found: www.ssp.sh/brain/markdo... Which one do you prefer and why?

submitted 11 days ago • 1 comment

I use (mostly) obsidian-project to manage my business. Now this function comes built in with Obsidian 1.9.0 with "Bases". Main advantage, the metadata of project-plugin was hidden in setting folders. Bases uses a YAML with the extension `.base` that gets integrated as normal file. Nice!

submitted 12 days ago • 0 comments

TIL—I can run DuckDB queries within #neovim via MCP connected to my remote MotherDuck databases 🤯. 📺 asciinema.org/a/720375 🧩 Plugin: MCPHub.nvim & Avante.nvim integration The internet is amazing!

submitted 13 days ago • 2 comments

Besides the hype of AI, I still think GenBI is one of the best places to use it (e.g., self-serving BI). Below are two practical examples: ClickHouse's real-life interaction and query database and generating dashboards/models with LLMs in Rill. One key element, though, is speed.

submitted 13 days ago • 1 comment

Obsidian, you beauty.

submitted 14 days ago • 0 comments

I love the sentiment of «Open standards over silos». It reminds me of the @stephango.com philosophy for Obsidian: «File over app». With open data platforms, open lakehouse, composable data stacks, or declarative data stacks, this is possible. We just need to figure out the terminology.

submitted 14 days ago • 0 comments

What would you call this stack? An "open format", "open data stack architecture", or what else? Or would you say this is the lakehouse architecture? And also, what's missing in the architecture?

submitted 15 days ago • 4 comments