Hacker News: joelschw

New comment by joelschw in "Model Training as Code"

joelschw — Mon, 29 Jun 2026 08:51:44 +0000

No this is a composability surface, Flyte is their workflow engine for durable execution etc

New comment by joelschw in "Ask HN: Who is hiring? (April 2026)"

joelschw — Wed, 01 Apr 2026 16:43:39 +0000

PhysicsX | Principal AI Engineer | London / NY / Remote | Full time | physicsx.ai

We're building an AI-native platform that turns physics simulations from days-long jobs into millisecond inference. Think: training deep learning surrogates on CFD, FEA, thermal data, and Large Physics Model (LPM) frontier research so engineers can explore thousands of design variants in a single session — across aerospace, semiconductors, energy, and automotive.

You should have the scar tissue of running agents in production and strong opinions on what the future looks like.

We're ~170 people, raised $135M Series B, and backed by General Catalyst, Atomico, Siemens, and Applied Materials. The team includes people from F1, SpaceX, Google, Palantir, and Tesla.

If this sounds interesting: https://job-boards.eu.greenhouse.io/physicsx/jobs/4804769101

Or email me directly: joel.schwarzmann [at] physicsx [dot] ai

New comment by joelschw in "Show HN: I built a tool to un-dumb Claude Code's CLI output (Local Log Viewer)"

joelschw — Tue, 17 Feb 2026 11:13:56 +0000

You could also use braintrust here https://github.com/braintrustdata/braintrust-claude-plugin

New comment by joelschw in "BBC director general and News CEO resign in bias controversy"

joelschw — Mon, 10 Nov 2025 09:14:43 +0000

January 6th was so egregious, I'm really not sure why they felt the need to exaggerate it.

New comment by joelschw in "Farewell Pandas, and thanks for all the fish"

joelschw — Thu, 29 Aug 2024 14:39:42 +0000

Huge fan of Ibis, the value isn't that you can now use DuckDB... it's that your syntax will work when the next cool thing arrives too

New comment by joelschw in "DuckDB in Action"

joelschw — Tue, 19 Mar 2024 10:27:42 +0000

Or you use Ibis and switch between the two at will!

New comment by joelschw in "Vizro – toolkit for creating modular data visualization applications"

joelschw — Wed, 27 Sep 2023 08:51:59 +0000

I'd put this in the readme!

New comment by joelschw in "Vizro – toolkit for creating modular data visualization applications"

joelschw — Wed, 27 Sep 2023 08:02:06 +0000

Big fan of evidence, really elegant design. There is also space for both declarative and imperative approaches when it comes to dashboarding / reporting etc.

Vizro – toolkit for creating modular data visualization applications

joelschw — Tue, 26 Sep 2023 17:15:33 +0000

Article URL: https://github.com/mckinsey/vizro

Comments URL: https://news.ycombinator.com/item?id=37662561

Points: 133

# Comments: 21

New comment by joelschw in "Git and Jupyter Notebooks Guide"

joelschw — Fri, 07 Jul 2023 13:36:50 +0000

The native GitHub feature in preview will make this a lot better for those able to use it https://github.blog/changelog/2023-03-01-feature-preview-ric...

New comment by joelschw in "SQL + M4 = Composable SQL"

joelschw — Wed, 28 Dec 2022 12:01:43 +0000

Whilst I hate it - Jinja is so much more accessible for SQL coders.

New comment by joelschw in "Useful Python decorators for data scientists"

joelschw — Mon, 23 May 2022 09:28:34 +0000

I'm not sure it's framed like that, it's specifically talking about production patterns

New comment by joelschw in "DoltLab: The Self-Hosted DoltHub"

joelschw — Tue, 01 Feb 2022 20:08:37 +0000

If git did it...

New comment by joelschw in "Kedro – Creating reproducible, maintainable and modular data science code"

joelschw — Sun, 23 Jan 2022 05:58:14 +0000

Some similar concepts, but Kedro is specialised on the ML engineering workflow and team collaboration

New comment by joelschw in "Kedro – Creating reproducible, maintainable and modular data science code"

joelschw — Sat, 22 Jan 2022 21:04:51 +0000

Here you go - https://kedro.readthedocs.io/en/0.17.6/09_development/03_com...

New comment by joelschw in "Kedro – Creating reproducible, maintainable and modular data science code"

joelschw — Sat, 22 Jan 2022 20:11:32 +0000

Will point to some experiences open source users have had on this journey

https://medium.com/hacking-talent/production-code-for-data-s...

https://medium.com/google-cloud/migrate-kedro-pipeline-on-ve...

New comment by joelschw in "Kedro – Creating reproducible, maintainable and modular data science code"

joelschw — Sat, 22 Jan 2022 19:27:41 +0000

Some users have reported the Targets library is similar for that ecosystem

New comment by joelschw in "Kedro – Creating reproducible, maintainable and modular data science code"

joelschw — Sat, 22 Jan 2022 18:30:32 +0000

Kedro sort of fits into a niche where it just overlaps somewhat with 'orchestrators' like Prefect, Metaflow, Dagster, Airflow and others. What makes it slightly different is that it it is focused on the rapid development journey to production, providing guardrails for teams to co-develop ML projects in a way that nudges software engineering best practice and clean code.

The 'finished article' in many cases should be deployed in production in one of those tools which provide specialised bells and whistles like scheduling, monitoring and observability.

Regarding MLFlow, there is also a slight overlap in terms of experimentation, but not things like model serving. Kedro has a mechanism to track experiments, but it's more designed to give users with zero infrastructure something for free out of the box. It's been built in a way that it can be repurposed for more dedicated experiment tracking tools - the folks at neptune.ai built their own plug-in for this purpose: https://docs.neptune.ai/integrations-and-supported-tools/aut...

New comment by joelschw in "Kedro – Creating reproducible, maintainable and modular data science code"

joelschw — Sat, 22 Jan 2022 16:44:53 +0000

I should clarify by short term - the foreseeable future is a more appropriate description

New comment by joelschw in "Kedro – Creating reproducible, maintainable and modular data science code"

joelschw — Sat, 22 Jan 2022 16:35:00 +0000

Hello maintainer here - hopefully some of our open source users chime in, but my favourite will always be NASA: https://github.com/nasa/ML-airport-configuration

Donation is the normal term for proejcts joining the foundation - in doing so you establish a steering committee where members are entitled to voting rights. To graduate as an incubation project, 5 organisations need to join your board and the project is thus the priorities of the project will no longer be driven by just one organisation.

In the short term there is still a full time internal team staffed and maintaining the project, but excitingly we now have a mechanism for new collaborators to properly come onboard.