| | At Fidelity, your tech career isn’t a set path—it’s an open door. Fidelity provides the support, learning platforms, and coaching you need to build a long-term career. You can explore different areas of the company, pursue opportunities on different teams, and grow into leadership roles. Join a culture that prioritizes your growth as much as new technology, offering you the versatility to build a future that’s uniquely yours. Explore what’s possible in a tech career at Fidelity.
*Message from this week's sponsor, Fidelity Careers |
| | |
| | | | Local models used to be slow, awkward, and clearly behind frontier APIs. Vicki Boykis argues that's changed. This post walks through her current local agentic coding setup with LM Studio, Pi, Docker, and Gemma models, including what works, what still breaks, and why local AI is becoming much more practical for technical work.
Vicki Boykis |
| | |
| | Most data pipelines don’t fail because they need more exotic tooling. They fail because of fragile assumptions, premature optimization, and complexity that leaks into downstream analysis. This post walks through a simple daily-batch pipeline and shows how raw, intermediate, and analytics layers can make the system easier to debug, reproduce, and trust.
Practical Data Community | Bruno Masciarelli |
| | |
| | Quarto is a publishing system for computational work that makes it easy to reuse the same source across reports, websites, PDFs, notebooks, and other formats. This post explores why Quarto has become so widely used, from multi-language support and cleaner defaults to Typst-powered PDFs and an extension system that makes Quarto easy to customize.
Mickaël Canouil |
| | |
| | AI is speeding up analytics work, but the fundamentals still decide whether anyone trusts the output. dbt’s 2026 State of Analytics Engineering Report looks at AI-assisted coding, governance gaps, infrastructure costs, and the growing pressure to deliver reliable insights faster. Learn more.
// sponsored |
| | |
| | | | Here's a practical, candid guide to the modern tech interview loop, seen through an ML lens. Silvia Sapora covers startup vs big-company hiring, prep strategy, ML learning resources, mock interviews, negotiation, and the lessons she’d carry into the next search.
Silvia Sapora |
| | |
| | | BigSet turns plain-English dataset ideas into structured tables. Describe what you want, and it infers a schema, researches public web sources, verifies and deduplicates rows, then exports CSV or XLSX. It’s experimental, but the idea here is compelling: lightweight, refreshable datasets without hand-building a scraper first.
GitHub | TinyFish |
| | |
| | SQL schemas are easier to understand when you can see the relationships. This open-source browser tool turns CREATE TABLE statements into an interactive ER diagram, with local-only processing, relationship highlighting, and inline edits that update the SQL. The repo describes it, and here’s a live example.
SQL to ER Diagram | Royal Bhati |
| | |
| | | Since the original Papers with Code went away, there’s been a gap for tracking AI papers, code, methods, benchmarks, leaderboards, conferences, and related artifacts in one place. This new version is still early, but it looks promising for keeping track of fast-moving ML research.
Curated daily from arXiv and Hugging Face |
| | |
| | | Mohand Alrasheed found a seashell in the middle of the desert and wanted to know why. From there, he turns shell identification into a shape-analysis problem, using contour extraction, shape normalization, PCA, and latent-space mapping to compare shell forms. There were probably easier ways to get an answer, but the detour is the fun part.
Mohand Alrasheed |
| | |
| | | This week's issue sponsored by: | | |
|
