What are you working on? Any new ideas which you're thinking about?
AI Scheduling Agent. See 1-min demo: https://www.youtube.com/watch?v=Pu5RQAGOaG4
Would love to know what you think.
Want to test it out? Sign up to the waitlist at https://brice.ai and I'll give you access tomorrow.
Integrating my time series database (https://github.com/dicroce/nanots) as the underlying storage engine in my video surveillance system, and the performance is glorious. Next up I'm trying to decide between a mobile app or AI... and if AI local or in the cloud?
Working on an manga comic creator called manga me: www.manga-me.me
What's different about it is that we've figured out character consistency with AI generated images, as well as text legibility. Most AI models don't do small text very well, and don't do consistent characters. We've tried to fix that.
Working on a programming language: https://antilisp.com, a Lisp used for code generation in other languages.
The language is heavily inspired by Python for the dev UX, and the interpreter is written in RPython (what Pypy uses). Rewriting to RPython was tedious, but the 80x speedup was worth it.
I'm building a catalog for health care price transparency data that aggregates the rates published by all insurers, to put everything in one place and make it easier for developers / researchers to access this data. https://www.accessmrf.com/
Website proofreading and auditing tool: https://www.triplechecker.com
Idea born out of my own frustration at finding typos at my prior company. I wanted a tool to crawl my website daily and uncover new errors. That’s how TripleChecker was born.
Another Chart - Web Component chart library
Wanted to try out vibe coding, to see how far it could take me.. pretty far it seems.. Just a small web component to display charts, supports line and bar chart for now.
We're building https://actory.ai. QA is necessary but sucks, LLMs are awesome, "vibe coding" is very cool, so ... we're building "vibe QA!"
https://finbug.xyz/, free software tools for Finnish language learners continues to be my primary project, in between long bouts of Anki cards. I recently revamped and standardized the CSS a little among the various online tools, and I quite like how they look now.
I'm dealing with mime types and max file sizes for an uploader, and improving error messages. Instead of relying on the file name to detect the mime type, I'm using the file binary header instead to reject dodgy files (for ex a `sample.jpg` file that is actually a ZIP or EXE under the hood).
Side project to enable speech-to-text functionality on my Ubuntu laptop so I can prompt AI coding tools fast:
I’m building a custom vocalist/DSP AI. I’ve never built any kind of neural net beyond a toy demo, but I’ve been programming for ~25 odd years.
Think like ACE Studio, but I’m going much less for pitch performance and much more for clarity, expressiveness and human realism.
Very much at the data labeling phase but a little bit beyond the crude initial experiment phase.
I'm building a tool that would let one generate Spring Boot integration tests based on Postman collections.
I think there is a gap between exploratory testing and more structured forms of testing. So I am trying to make a tool for that for myself. If I like the outcome I'll open-source it.
Anonymization of PII data in documents using diffusion models - I'm in the process of reproducing academic papers. The idea is you can replace sensitive information from financial/medical documents with synthetic analogues without visually altering them, so they can be kept/used for AI training
Preparing a new country (Switzerland) for a tiny daily birdwatching game I've been running with my gf for a few years in collaboration with a local organisation.
And at the same time working on getting the first play testing version ready for our new geo-location based game also about birdwatching.
i am working on GenAI email-based service in my spare time. It is still in its infancy but let me explore LLama3 feature on Ollama. I plan to open source its core if no similar ideas are out of there.
My Misterio docker based tool is searching new feature request... https://github.com/daitangio/misterio/
Also, I am playing a bit with Zulip Chat, which I find quite well done and easy to self-host, considering its complexity: https://github.com/zulip/docker-zulip
Last but not least, I suggest a new Murderbook novel... https://amzn.to/3TMJdlh because there is not only coding!
subscreen.app — working on an app that synchronizes with the movie/show you watch and displays subtitles on your phone (in the original language or a language of your choice).
Tech stack: - Python + opensubtitles.org for the data pipeline - Whisper for speech recognition - React Native for the mobile client
Current state: tech demo. The app works fine and already helps a lot — for me and my wife (both non-native English speakers) it makes watching movies in Dutch cinemas much easier, by showing English subtitles on our phones instead of the Dutch ones provided by the theater.
The biggest issue now is subtitle quality and legal status. Opensubtitles provides a lot of data, but the quality is often questionable, and the legal status is rather gray/black.
Any legal or data-related advice would be appreciated!
Tritium: https://tritium.legal (web preview: https://tritium.legal/preview)
Tritium is the legal integrated drafting environment: an egui Rust project to bring the IDE to corporate lawyers.
I'm working on Botnet of Ares [0], a hacking roguelike set in a cyberpunk world where everything is connected.
[0] https://store.steampowered.com/app/3627290/Botnet_of_Ares/
Building an AS/RS for trading cards. I did my POC smaller scale hacked together and now I'm building v1 (which I'm having to fight second system syndrome pretty hard on.) After getting very refactoring reluctant with untyped python, I'm making the transition to Rust and enjoying it quite a bit.
Launched a couple of weeks ago:
Vibe Interview simulates real job interviews using AI. Master every interview stage, from recruiter to technical rounds. Reply and I'll give you free minutes for call simulations.
A competitive rock-paper-scissors game. Something where replay-ability is high and low barrier to entry. I have completed most of the core features that I wanted and now I'm fine-tuning things, fixing bugs, etc...:
I built an internal monitoring tool that tracks new blockchain software releases and made it publicly available as a website. It's particularly useful for people who run and maintain blockchain nodes: http://chainrelease.info
I've still working on
It's a Google Meet attendance & chat tracker, and it's starting to pick up a bit. A few teachers & other people are using it and enjoying it which is really awesome!
Since I posted about https://github.com/theopfr/somo on HN it got a lot of attention and many new contributors! I'm working through lots of PRs and hoping to release a new version soon with all the new features :)
I'm working at BundleJoy: a Shopify app for creating boxes (bundles) with a better CX and some nice features (quantity rules, packaging options, collect customer input, etc).
AppStore: https://apps.shopify.com/bundlejoy
Still working on dédédé [1] - it's a simple web-based platform to share the "good, bad, and why"s of urban spaces. We're slowly adding functionalities and crushing bugs, an iOS app is in the pipeline too!
Finally had a project where I could use Go in, I built a small tool that runs a set of queries in Gitlab's code search, puts the result in a database, and generates a chart comparing two search results; the objective is to monitor adoption of a component library / design system over time.
I'm working on getting all my supported iOS/MacOS/WatchOS/TVOS apps ready for Version 26 (Liquid Glass).
It introduces quite a few changes. In my shipping apps, I'll probably be simply telling the OS not to use Liquid Glass (for now), but for my various test harni, I will need to adapt. Looks like a fair bit of work.
A free site to find sports/games around you (tennis, basket, chess, etc). You propose a game at a certain date/place, and people around can reach out to join. Let me know if you want to try it out (ready in the coming days).
Sharpee: A Typescript based Interactive Fiction platform that uses event sourcing and a post turn text service to emit updates.
Architecture uses Traits (data) and Behaviors (logic) to implement things in the world model.
I built https://founderfodder.com as an alternate way to read HN stories, plus it has summaries of the stories and the comments if you want to save some time
I just built out a MCP for ShareSEER https://shareseer.com/claude to enable the data that I have about US public company filings, financials & insider transaction right within Claude
https://Podskim.com is a way to skim through podcasts like a TikTok sans the brain rot. It also has some fact checking and topic monitoring behind the scenes. Haven't figured out a business model for it yet but has been fun to keep poking it
StarStories: You get a real physical book featuring members of your family
1. Upload photos of your family members (or describe them if you don't want to upload)
2. Select a topic
3. See draft book
4. Make edits if you want
5. Order book
6. Read book to your kids
7. Read book to your kids
8. Continue on loop
Refactoring the front end of my Java web app to follow basic principles of algebraic data types.
The goal is to make the code better organized, easier to read, maintain and extend.
Working on https://github.com/netdur/llama_cpp_dart it is llama.cpp binding for Dart first then Flutter I am currently working on multimodal support, add vision and working on audio
Excel formula to postgres SQL compiler. Reminiscent of Salesforce formulas.
Demo uses postgres compiled for WASM so demo runs on an actual postgres db.
I'm building a FOSS educational platform similar to Khan Academy but with open educational resources like MIT OCW. I'm also planning on hosting the videos on my PeerTube instance and creating a torrent for Academic Torrents.
I'm working on EyesOff[1] - v2.0.0 should be out soon.
It's a simple (currently macOS) application which aims to target shoulder surfing by using a locally running neural network to detect those looking at your screen.
Solo building this project: https://www.usesubwise.com/
A side project that started out of personal frustration: I kept getting charged for subscriptions I forgot I even signed up for.
Subwise helps you track all your recurring payments across platforms — with smart reminders, spending insights, and cost optimization tips (like switching to annual billing or downgrading tiers you don’t use). Think of it as Mint meets a subscription-focused dashboard, with a browser extension that auto-detects subscriptions during checkout.
The bigger picture: I want to reduce "subscription fatigue" and help users stay financially aware, especially as more services go SaaS/paywalled. Eventually thinking about small biz or team-level use cases too.
Tech-wise: Built with Next.js + TypeScript, Postgres + Drizzle ORM, and a lightweight Chrome extension. Stripe is integrated for premium features. Everything’s still MVP, but slowly improving!
Would love feedback if anyone here manages a bunch of subscriptions (or forgets them like I do).
I am in the early design phases for a local-first software forge, intended for unreliable networks. I should be able to continue tinker open source software, and limited collaboration, even if network access is unreliable.
This is starting to overlap with building a tool server for personal AI agents.
Real-time synthetic data generation with built in connectors https://github.com/glassflow/glassgen Next step is to extend it as a server module so you can run it remotely
I'm bootstrapping a [Bluesky analytics, Bluesky+X+Mastodon post scheduling tool called TheBlue.social](https://theblue.social)
But working on it for past 7 months. It's running and I'm tweak/adding features while marketing it.
gemini-copilot: https://github.com/binora/gemini-copilot
Just trying to make gemini-cli work with my enterprise copilot subscription.
Started as a curiosity and turned into a small little app (not encouraging or discouraging use) https://github.com/issafram/torrent-ratio-booster/tree/main
Tiny program that batch renumbers sections/clauses in documents.
It is like MS Word "Bullets and numbering" but it's a small UNIX filter, no GUI, much faster and smoother than MS Word or Google Docs.
Perhaps the beginning of a markup language for text or HTML files intended to be converted to MS Word.
1) A website to measure and detect coil whine. It's been bugging me on my new Dell screen, but Dell says "it's within specs". 2) An AI-generated artwork platform with open firmware for Eink frames. 3) Server Radar: https://radar.iodev.org