| | Nebius Token Factory is designed for teams going beyond APIs. Deploy and scale open-source LLMs, fine-tune them to your use case, and manage access, data, and cost as your product grows. Everything you need to build serious AI, without stitching together tools. From LLM to production system, in one platform →
*Message from this week's sponsor, Nebius Token Factory |
| | |
| | 🔗 Note: if you don’t see links in this email, it’s also available on the web: dataelixir.com/archive/issue-568.html |
| | |
| | Most data work assumes exact answers are the goal. At scale, that assumption breaks. This is a clear introduction to probabilistic data structures and when it makes sense to trade precision for speed.
Kirill Bobrov |
| | |
| | Stop copying full tables and start tracking what changed instead. This post introduces change data capture, then compares three different approaches and shows how they work and where each one breaks.
Podo Stack | Ilia Gusev |
| | |
| | | Foundation models for time series sound great until you ask what “time series” actually means. Finance, weather, and physiology share almost nothing. This post digs into why that mismatch matters.
shako |
| | |
| | SQLite keeps picking up features people still associate with bigger systems, such as JSON queries, window functions, full-text search, strict tables, and more. This is a quick tour of what’s there, and a good reminder that “just SQLite” covers more ground than most people realize.
slicker |
| | |
| | Training your own language model isn’t just for big labs anymore. This walkthrough builds a transformer on 2.93 billion tokens of Victorian text, covering everything from dataset filtering to tokenizer design and renting GPUs on Vast AI.
Estragon | Trip |
| | |
| | | This started as a Stanford course and is now turning into an online handbook for textile-based data visualization. It builds on the Dear Data book by Giorgia Lupi and Stefanie Posavec, including examples, datasets, and patterns that make physical data work feel surprisingly systematic.
Quinn Daedal |
| | |
| | | Give it a job description and this Claude skill generates realistic interview questions, runs mock interviews and critiques your answers. It turns interview prep into a full feedback loop.
Noam Segal |
| | |
| | | This is a great demo, and it doubles as a sharp reminder that SQL can model and render structured systems in ways most people never explore. It’s a full, playable chessboard. No frontend. Just SQL.
DB Pro | Jay |
| | |
| | | This week's issue sponsored by: | | |
|
