Ask HN: What are you working on? (March 2025) • Hacker News Board

Re: Ask HN: What are you working on? (March 2025)

#911

Post by sharpshadow » Tue, Apr 01, 2025, 12:47 PM UTC

andrethegiant wrote: post #97

I'm working on pure.md[1], which lets your scripts, APIs, apps, agents, etc reliably access web content in markdown format. Simply prefix any URL with `pure.md/` and you get the unblocked markdown content of that webpage. It avoids bot detection and renders JavaScript-heavy websites, and can convert HTML, PDFs, images, and more into pure markdown. pure.md acts as a global caching layer between LLMs and web content. I…

Works great on mobile thanks, helpful tool to bypass flaky websites, js and even some paywalls.

Re: Ask HN: What are you working on? (March 2025)

#912

Post by teleforce » Tue, Apr 01, 2025, 1:10 PM UTC

enisberk wrote: post #443

Finishing up my PhD thesis on low-resource audio classification for ecoacoustics. Our partners deployed 98 recorders in remote Arctic/sub-Arctic regions, collecting a massive (~19.5 years) dataset to monitor wildlife and human noise. Labeled data is the bottleneck, so my work focuses on getting good results with less data. Key parts: - Created EDANSA [1], the first public dataset of its kind from these areas, using a…

Hi Enis, it seems a very interesting project. I myself with my team are currently working with the non-stationary of physiological and earthquake seismic public data mainly based on the time-frequency distributions, and the results are very promising.

Just wondering if the raw data that you've mentioned are available publicly so we can test our techniques on them or they're only available through research collaborations. Either way very much interested on the potential use of our techniques for the polar research in Arctic and/or Antarctica.

Re: Ask HN: What are you working on? (March 2025)

#913

Post by sircastor » Tue, Apr 01, 2025, 2:17 PM UTC

999900000999 wrote: post #115

FOSS MTG inspired digital card game. I love card games, but for digital card games the business model is beyond predatory. If you need a specific card your option is to basically buy a pack. Let’s say this is about 3$ give or take. But if it’s a specific rare card, you can open a dozen of so packs and still not get the specific card you want. This can go on indefinitely, and apologists will claim you can just work ar…

Many years ago Decipher (who made the Star Trek and Star Wars TCGs) rolled out a web platform for playing their games. It was the business model but with none of the advantages of the physical property. You would spend money on their platform to buy their digital cards, to play only there, and when you left the cards just disappeared into the void.

Re: Ask HN: What are you working on? (March 2025)

#914

Post by mhdi_kr99 » Tue, Apr 01, 2025, 2:41 PM UTC

I'm working on a GenAI learning platform (focusing on software engineering). I want it to not suck, not to be another LLM wrapper. I will make it public as soon as I find it useful myself.

Re: Ask HN: What are you working on? (March 2025)

#915

Post by 999900000999 » Tue, Apr 01, 2025, 2:54 PM UTC

enisberk wrote: post #443

Finishing up my PhD thesis on low-resource audio classification for ecoacoustics. Our partners deployed 98 recorders in remote Arctic/sub-Arctic regions, collecting a massive (~19.5 years) dataset to monitor wildlife and human noise. Labeled data is the bottleneck, so my work focuses on getting good results with less data. Key parts: - Created EDANSA [1], the first public dataset of its kind from these areas, using a…

How do I download the sounds, seems like a great resource for game developers and other artists

Re: Ask HN: What are you working on? (March 2025)

#916

Post by wfn » Tue, Apr 01, 2025, 4:46 PM UTC

erekp wrote: post #730

andrethegiant wrote: post #97

I'm working on pure.md[1], which lets your scripts, APIs, apps, agents, etc reliably access web content in markdown format. Simply prefix any URL with `pure.md/` and you get the unblocked markdown content of that webpage. It avoids bot detection and renders JavaScript-heavy websites, and can convert HTML, PDFs, images, and more into pure markdown. pure.md acts as a global caching layer between LLMs and web content. I…

how do you exactly fallback to common crawl? isn't the cost to even hold and query common crawl insane?

Since I was just looking at this accidentally, here are some examples of how to query at a ~cent-per-query cost level (just examples but quite illustrative): https://commoncrawl.org/blog/index-to-warc-files-and-urls-in...

Re: Ask HN: What are you working on? (March 2025)

#917

Post by jlcases » Tue, Apr 01, 2025, 5:44 PM UTC

I've been working on a documentation system that improves how AI assistants understand project context. The problem I'm solving is that traditional documentation approaches are too fragmented for AI to effectively process.

After struggling with vague outputs from AI coding assistants, I started experimenting with MECE principles (Mutually Exclusive, Collectively Exhaustive) to organize technical documentation hierarchically. Think of it as creating a "mental model" for AI rather than just reference docs.

Some early results: - 40% reduction in back-and-forth iterations with AI assistants - Much more consistent code style and architectural patterns - Better preservation of domain knowledge across the project

Currently refining the metadata structure and creating templates for different project types. The system works especially well with Cursor AI and similar assistants that can process structured context.

If anyone else is exploring this space of "AI-optimized documentation," I'd love to exchange notes.

Re: Ask HN: What are you working on? (March 2025)

#918

Post by kelseyfrog » Tue, Apr 01, 2025, 5:51 PM UTC

While reading AI Engineering[1], I was inspired to take a crack at an idea I've had for a while - diffusion-based LLM. The accessibility of shakespeare.txt[2] and the proliferation of tinyGPT implementations has made vibe-coding LLM research on the brink of possibility. It does help to have a familiarity in the datascience/ML/AI space in order to check and guide appropriately, but it's amazing how well the virtuous cycle is beginning to work.

1. https://www.oreilly.com/library/view/ai-engineering/97810981...

2. https://gist.github.com/blakesanie/dde3a2b7e698f52f389532b4b...

Re: Ask HN: What are you working on? (March 2025)

#919

Post by bbkane » Tue, Apr 01, 2025, 6:45 PM UTC

dvcrn wrote: post #769

A buddy and me have been working on https://microfn.dev for a while: A platform for creating, managing and composing tiny (micro) javascript functions and using them from different places like webhooks, cron, MCP or AI agents. It's still very heavy in development but the gist is: Say you have a cool idea for something small you want to automate or run - instead of thinking about hosting, workers, lambdas and what not…

Sounds neat, but please make testing first class - it should be easy to run functions locally, or to have the same code "infrastructure" run in dev/test/prod environments, etc

Re: Ask HN: What are you working on? (March 2025)

#920

Post by enisberk » Tue, Apr 01, 2025, 6:57 PM UTC

teleforce wrote: post #912

enisberk wrote: post #443

Finishing up my PhD thesis on low-resource audio classification for ecoacoustics. Our partners deployed 98 recorders in remote Arctic/sub-Arctic regions, collecting a massive (~19.5 years) dataset to monitor wildlife and human noise. Labeled data is the bottleneck, so my work focuses on getting good results with less data. Key parts: - Created EDANSA [1], the first public dataset of its kind from these areas, using a…

Hi Enis, it seems a very interesting project. I myself with my team are currently working with the non-stationary of physiological and earthquake seismic public data mainly based on the time-frequency distributions, and the results are very promising. Just wondering if the raw data that you've mentioned are available publicly so we can test our techniques on them or they're only available through research collaborati…

Hi teleforce, thanks! Your project sounds very interesting as well.

That actually reminds me, at one point, a researcher suggested looking into geophone or fiber optic Distributed Acoustic Sensing (DAS) data that oil companies sometimes collect in Alaska, potentially for tracking animal movements or impacts, but I never got the chance to follow up. Connecting seismic activity data (like yours) with potential effects on animal vocalizations or behaviour observed in acoustic recordings would be an interesting research direction!

Regarding data access:

Our labeled dataset (EDANSA, focused on specific sound events) is public here: https://zenodo.org/records/6824272. We will be releasing an updated version with more samples soon.

We are also actively working on releasing the raw, continuous audio recordings. These will eventually be published via the Arctic Data Center (arcticdata.io). If you'd like, feel free to send me an email (address should be in my profile), and I can ping you when that happens.

Separately, we have an open-source model (with updates coming) trained on EDANSA for predicting various animal sounds and human-generated noise. Let me know if you'd ever be interested in discussing whether running that model on other types of non-stationary sound data you might have access to could be useful or yield interesting comparisons.