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

As a bio hobbyist, this is fantastic! I don't do enough volume of PCR to think of it as expensive, but your use case of high-volume/automatic sounds fantastic! (And so many other types of reagents and equipment are very expensive).

Some things that would be cool

  - Along your lines: In general, cheap automated setups for PCR and gels
  - Cheap/automatic quantifiable gels. E.g. without needing a kV supply capillary, expensive QPCR machines etc.
  - Cheaper enzymes in general
  - More options for -80 freezers
  - Cheaper/more automated DNA quantification. I got a v1 Quibit which gets the job done, but new ones are very expensive, and reagent costs add up.
  - Cheaper shaking incubator options. You can get cheap shakers and baters, but not cheap combined ones... which you need for pretty much everything. Placing one in the other can work, but is sub-optimal due to size and power-cord considerations.
  - More centrifuges that can do 10kG... this is the minimum for many protocols.
  - Ability to buy pure ethanol without outrageous prices or hazardous shipping fees.

- Not sure if this is feasible but... reasonable cost machines to synthesize oglios?

Currently working on https://tinyZKP.com

It's an API that allows zero-knowledge proofs to be generated in a streaming fashion, meaning ZKPs that use way less RAM than normal.

The goal is to let people create ZKPs of any size on any device. ZKPs are very cool but have struggled to gain adoption due to the memory requirements. You usually need to pay for specialized hardware or massive server costs. Hoping to help fix the problem for devs

Working on SecureFlow (https://codepathfinder.dev/secureflow-ai/) - think of claude-code style but for hunting security vulnerabilities.

The goal is to catch vulnerabilities early in the SDLC by running agentic loop that autonomously hunt for security issues in codebases.Currently available as a CLI tool, VSCode extension.I've been actively using to scan WordPress, odoo plugins and found several privilege escalation vuln. I have documented as blog post here: https://codepathfinder.dev/blog/introducing-secureflow-cli-t...

I wanted to build my own speech-to-text transcription program [1] for Discord, similar to how zoom or google hangouts works. I built it so that I can record my group's DND sessions and build applications / tools for VTTs (Virtual TableTop gaming).

It can process a set of 3-hour audio files in ~20 mins.

I recorded a demo video of how it works here: https://www.youtube.com/watch?v=v0KZGyJARts&t=300s

[1] https://github.com/naveedn/audio-transcriber

I alluded to building this tool on a previous HN thread: https://news.ycombinator.com/item?id=45338694

I'm building an open source NAT traversal and networking framework called P2PD. Built from the ground up to allow things like multi-network interface applications, improved network programming in Python, and if people want it: an easy way to bypass NATs. The thing is: it depends on public servers for some of this which tends to change a lot, causing errors when they're all down.

What I'm building at the moment is a server monitoring solution for STUN, TURN, MQTT, and NTP servers. I wanted to allow the software for this to be portable. So I wrote a simple work queue myself. Python doesn't have linked-lists which is the data structure I'm using for the queues. They allow for O(1) deletes which you can't really get on many Python data structures. Important for work items when you're moving work between queues.

For the actual workers I keep things very simple. I make like 100 independent Python processes each with an event loop. This uses up a crap load of memory but the advantage is that you can parallel execution without any complexity. It would be extremely complex trying to do that with code alone and asyncio's event loop doesn't play well with parallelism. So you really only want one per process.

Result: simple, portable Python code that can easily manage monitoring hundreds of servers (sorry didnt mean for that to sound like chatgpt, lmao, incidental.) The DB for this is memory-based to avoid locking issues. I did use sqlite at first but even with optimizations there were locking issues. Now, I only use sqlite for import / export (checksums.)

Not anything special by HN standards but work is here: https://github.com/robertsdotpm/p2pd_server_monitor

I'm at the stage now where I'm adding all the servers to monitor to it. So fun times.