Hacker News: pplonski86

New comment by pplonski86 in "Show HN: Which AI model is best for real data analysis?"

pplonski86 — Tue, 14 Apr 2026 12:35:37 +0000

We built a benchmark to evaluate LLMs on real data analysis workflows. Instead of single prompts, each task is a sequence of prompts (steps). It is similar to how a human data analyst works in practice. Each run is saved as full python notebook, including prompts, code and outputs. We evaluated runs across task completion, code correctess, output quality, reasoning and reliability. Each workflow is execuuted multiple times and scored automatically.

Modern LLMs perform very well on individual steps. The benchmark currently inludes 23 workflows from different data analysis tasks (EDA, ML, NLP, statistics ...). The top-3 models across the 23 workflows, gpt-oss:120b scored 9.87/10, followed by gpt-5.4 at 9.65/10, glm-5.1 at 9.48/10. Which is very high in my opinion. The results show that modern LLMs perform very well on data analysis tasks. All feedback is welcome! I uploaded all notebooks for each model https://github.com/pplonski/ai-for-data-analysis

Show HN: Which AI model is best for real data analysis?

pplonski86 — Tue, 14 Apr 2026 12:27:56 +0000

Article URL: https://mljar.com/analysis/

Comments URL: https://news.ycombinator.com/item?id=47764735

Points: 2

# Comments: 1

Ask HN: Im back end engineer, not front end – is this just excuse?

pplonski86 — Sat, 11 Apr 2026 16:20:12 +0000

When I hear: I'm not good at frontend, I'm good at backend - I'm starting to think this is not true.

If you don't care about the user in the UI, you probably don't care about the user in the backend either. Frontend problems are just easier to see — bad UI, confusing views, ugly design. Everyone notices. Backend problems are hidden — slow API response, bad errors, hard to manage code. It takes longer to notice, but users still suffer. So maybe it's not about frontend vs backend. Maybe it is just about caring (or not caring) about the end user. Have you seen this?

Comments URL: https://news.ycombinator.com/item?id=47731814

Points: 3

# Comments: 4

New comment by pplonski86 in "Claude Code's source code has been leaked via a map file in their NPM registry"

pplonski86 — Tue, 31 Mar 2026 14:02:07 +0000

I thought it was open source project on github? https://github.com/anthropics/claude-code no?

New comment by pplonski86 in "Google's 200M-parameter time-series foundation model with 16k context"

pplonski86 — Tue, 31 Mar 2026 08:40:56 +0000

Can someone explain ELI5 how it does work? and how many data points it can read?

Machine Learning for Humans and Now for LLMs

pplonski86 — Fri, 27 Mar 2026 14:22:11 +0000

Article URL: https://mljar.com/blog/structured-automl-reports-python-llm/

Comments URL: https://news.ycombinator.com/item?id=47542991

Points: 2

# Comments: 0

Show HN: Prompts Directory for Data Analyst

pplonski86 — Tue, 24 Mar 2026 07:37:11 +0000

Article URL: https://mljar.com/ai-prompts/data-analyst/

Comments URL: https://news.ycombinator.com/item?id=47499593

Points: 1

# Comments: 0

AutoResearch by Karpathy and the Future of Autonomous AI Research

pplonski86 — Mon, 23 Mar 2026 08:07:49 +0000

Article URL: https://mljar.com/blog/autoresearch-karpathy-autonomous-ai-research/

Comments URL: https://news.ycombinator.com/item?id=47486602

Points: 2

# Comments: 0

Show HN: AI Prompts for Data Professionals

pplonski86 — Mon, 23 Mar 2026 07:42:51 +0000

Article URL: https://mljar.com/ai-prompts/

Comments URL: https://news.ycombinator.com/item?id=47486460

Points: 1

# Comments: 0

AutoResearch by Karpathy and the Future of Autonomous Machine Learning

pplonski86 — Tue, 17 Mar 2026 16:26:39 +0000

Article URL: https://mljar.com/blog/autoresearch-karpathy-autonomous-ai-research/

Comments URL: https://news.ycombinator.com/item?id=47414882

Points: 2

# Comments: 0

GeoGPT – Chat-controlled GIS app built from a Jupyter Notebook

pplonski86 — Fri, 06 Feb 2026 11:40:41 +0000

Article URL: https://mljar.com/blog/chatgpt-gis-app-jupyter-notebook/

Comments URL: https://news.ycombinator.com/item?id=46911686

Points: 1

# Comments: 0

New comment by pplonski86 in "xAI joins SpaceX"

pplonski86 — Tue, 03 Feb 2026 12:27:29 +0000

Do we need rockets to put satelittes to the space? Cant it be done with baloons? https://www.youtube.com/watch?v=NFieAD5Gpms

New comment by pplonski86 in "The Codex App"

pplonski86 — Tue, 03 Feb 2026 12:20:09 +0000

It is not that easy to build such app from scratch ... it all requires a lot of work, even with AI help. I think the most important is to provide easy to use UI first, and if speed or some missing features will be blockers for further innovation step then maybe native app will be at some point created.

Build chatbot to talk with your PostgreSQL database using Python and local LLM

pplonski86 — Mon, 02 Feb 2026 11:17:24 +0000

Article URL: https://mljar.com/blog/chatbot-python-postgresql-local-llm/

Comments URL: https://news.ycombinator.com/item?id=46854722

Points: 1

# Comments: 0

New comment by pplonski86 in "Microsoft forced me to switch to Linux"

pplonski86 — Wed, 28 Jan 2026 21:10:33 +0000

I have dual boot on decent laptop, doing nothing, on windows fan is always on, computing something? On Linux it is just silent

New comment by pplonski86 in "The browser is the sandbox"

pplonski86 — Tue, 27 Jan 2026 08:16:58 +0000

Thank you! With WASM I can’t use all pypi packages and can’t connect to database, that’s why I’m looking for python based solution

New comment by pplonski86 in "Kimi Released Kimi K2.5, Open-Source Visual SOTA-Agentic Model"

pplonski86 — Tue, 27 Jan 2026 08:14:45 +0000

Thank you! Exactly what I was looking for

New comment by pplonski86 in "AI code and software craft"

pplonski86 — Tue, 27 Jan 2026 07:28:09 +0000

What if AI starts to have sense of craft? we just miss the verify and critique models, that will tell other models what looks good

New comment by pplonski86 in "ChatGPT Containers can now run bash, pip/npm install packages and download files"

pplonski86 — Tue, 27 Jan 2026 07:26:07 +0000

thank you for sharing, is there a new container for each code run, or it stays the same for whole conversation?

New comment by pplonski86 in "Kimi Released Kimi K2.5, Open-Source Visual SOTA-Agentic Model"

pplonski86 — Tue, 27 Jan 2026 07:22:59 +0000

There are so many models, is there any website with list of all of them and comparison of performance on different tasks?