frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

OpenAI's In-House Data Agent

https://openai.com/index/inside-our-in-house-data-agent
23•meetpateltech•1h ago

Comments

0xferruccio•58m ago
At Amplitude we built Moda which is super similar to this.

Our chief engineer Wade gave an awesome demo to Claire Vo some months back here: https://www.youtube.com/watch?v=9Q9Yrj2RTkg

I use this basically every day asking all sorts of questions

sjsishah•38m ago
Given my personal experience with various BI systems I think an AI agent like this is the perfect use case. These systems are operating on multiple layers of being wrong as is - layer 1 being your query is likely wrong, layer 2 being how you interpret the data is likely wrong.

Mix them together and you’re already deep in make believe land, so letting AI take over step 1 seems like a perfect fit.

I was hoping to read this article and be surprised by how OpenAI was able to solve the reliability problem, but alas.

htrp•12m ago
data problems are not tech problems but rather org problems
maxchehab•10m ago
Trust is the hardest part to scale here.

We're building something similar and found that no matter how good the agent loop is, you still need "canonical metrics" that are human-curated. Otherwise non-technical users (marketing, product managers) are playing a guessing game with high-stakes decisions, and they can't verify the SQL themselves.

Our approach: 1. We control the data pipeline and work with a discrete set of data sources where schemas are consistent across customers 2. We benchmark extensively so the agent uses a verified metric when one exists, falls back to raw SQL when it doesn't, and captures those gaps as "opportunities" for human review

Over time, most queries hit canonical metrics. The agent becomes less of a SQL generator and more of a smart router from user intent -> verified metric.

The "Moving fast without breaking trust" section resonates, their eval system with golden SQL is essentially the same insight: you need ground truth to catch drift.

Wrote about the tradeoffs here: https://www.graphed.com/blog/update-2

spiderfarmer•6m ago
I'm more interested in Kimi's In-House Data Agent

Open-Slopware

https://codeberg.org/small-hack/open-slopware
1•gpi•23s ago•0 comments

On this Day...1776 – January 1: The Flag [video]

https://www.youtube.com/watch?v=sV52AUVGc6I
1•mellosouls•1m ago•0 comments

Cognition Devin Review

https://app.devin.ai/review
2•lord_sudo•1m ago•0 comments

AltStore creators introduce CSAM Store Checker app

https://www.patreon.com/posts/introducing-csam-149431432
1•_han•3m ago•0 comments

Sicherheitslücke – gesperrte Bezahlmethoden trotzdem nutzbar bei smartsteuer.de

https://anton.dachauer.org/2026-01-27-smartsteuer.html
1•rizutato•4m ago•0 comments

LlamaBarn: A cosy home for your LLMs

https://github.com/ggml-org/LlamaBarn
1•tosh•7m ago•0 comments

I dont want AI that replace my taste, I want AI that help me use my taste better

https://emsh.cat/good-taste/
1•embedding-shape•9m ago•0 comments

Why the text terminal cursor is important for Accessibility

https://blind.guru/blog/2021-06-25-brick.html
1•lynx97•10m ago•0 comments

Show HN: Accurate LLM-based password guesser

https://github.com/Tzohar/PassLLM
2•Plarsy•13m ago•0 comments

Why "The AI Hallucinated" is the perfect legal defense

https://niyikiza.com/posts/hallucination-defense/
1•niyikiza•13m ago•1 comments

Creator Studio: Apple confuses with duplicate apps

https://www.heise.de/en/news/Creator-Studio-Apple-confuses-with-duplicate-apps-11158774.html
1•doener•13m ago•0 comments

"Mr. Burns" as a Software Engineer

https://github.com/arjun-krishna1/mrburns
1•arjun_krishna1•14m ago•0 comments

Show HN: A cross-framework Markdown/MDX parser to simplify content management

https://github.com/aymericzip/intlayer/blob/main/docs/docs/en/dictionary/markdown.md
2•intlayer_org•14m ago•0 comments

Clawdbot sheds skin to become Moltbot, can't slough off security issues

https://www.theregister.com/2026/01/27/clawdbot_moltbot_security_concerns/
1•thoughtpeddler•14m ago•0 comments

Crazy-Jumpman

https://codeberg.org/pgeorgi/crazy-jumpman/#english
1•todsacerdoti•14m ago•0 comments

1km tower in the desert is not progress it is a farewell letter to common sense

https://www.doubleglazinginbanbridge.co.uk/29-171405-1km-tower-in-the-desert/
2•patrakov•21m ago•0 comments

Google agrees to pay $68M to settle Google Assistant "false accepts" privacy ls

https://www.theguardian.com/technology/2026/jan/26/google-privacy-suit-settlement-voice-assistant
1•nuclearm•23m ago•1 comments

Lean 4.27.0

https://lean-lang.org/doc/reference/latest/releases/v4.27.0/
1•tzury•23m ago•0 comments

Show HN: o2go – a minimal, provider-agnostic OAuth2 client for Go

https://github.com/gokhanaltun/o2go
1•5gkhn2•24m ago•1 comments

Are AI cryptocurrencies a good investment in 2026?

https://altcoindesk.com/perspectives/learn/are-ai-cryptocurrencies-a-good-investment/article-22644/
1•aisshwarya20•24m ago•0 comments

Traditional Onboarding Excludes the People Tech Needs Most

https://vetswhocode.io/blogs/how-traditional-onboarding-excludes
1•bityard•24m ago•0 comments

MAME 0.285

https://www.mamedev.org/?p=559
1•chungy•24m ago•0 comments

UK-based pair behind messaging app accused of giving data to Iranian regime

https://www.theguardian.com/technology/2026/jan/29/iran-app-gap-messenger-tsit-user-data-uk-sussex
2•n1b0m•26m ago•0 comments

Benchmarking Reward Hack Detection in Code Environments via Contrastive Analysis

https://arxiv.org/abs/2601.20103
1•darshandesh1504•27m ago•1 comments

Sentry releases new CLI for developers and agents

https://cli.sentry.dev/
1•BYK•27m ago•0 comments

Flameshot

https://github.com/flameshot-org/flameshot
1•OsrsNeedsf2P•28m ago•0 comments

Next 50 Years of AI

https://www.dimamik.com/posts/next_50_years_of_ai/
2•dimamik•30m ago•0 comments

School is worse for kids than social media

https://unpublishablepapers.substack.com/p/school-is-way-worse-for-kids-than
1•amadeuspagel•31m ago•0 comments

NOAA Releases Updated Model of Earth's Geomagnetic Field

https://www.ncei.noaa.gov/news/upgraded-goemag-model-aids-energy-exploration
1•krunck•31m ago•1 comments

Ferrari vs. Markets

https://ferrari-imports.enigmatechnologies.dev/
3•merinid•33m ago•3 comments