A REST framework for FastAPI and SQLAlchemy, creatively called "fastapi-alchemy". A bit on the ambitious side, but I had the opportunity to open-source work I did for an employer.
It is running in production, but it needs extensive documentation and testing still before I'll package it and offer it to PYPI.
Ask HN: What are you working on? (May 2025)
561–570 of 1001 posts
Re: Ask HN: What are you working on? (May 2025)
#562A workout app that uses human readable/writable text file as input. I’m on the 4th iteration. It started as a bunch of spiral bound notebooks. I made an iOS prototype but realised it’s so niche that it would never get approved. Made a PWA that I’m currently using for about a year. Now I’m building a second version which I finally plan to share. The reason I’m making it is that I realized that any other style (mainly,…
Why use a separate app at all? I just use notepad (or its iOS equivalent).
Re: Ask HN: What are you working on? (May 2025)
#563CLI that abstracts rolling out multiple terminal-driven coding agents in parallel with the same prompt using git worktrees
Re: Ask HN: What are you working on? (May 2025)
#564I'm working on yet another spending analysis app. How will this one be different?
- Stores transactions locally in a SQLite database, so in theory you could build your own Front End if mine didn't do everything you wanted. Think someone who wants to do a fancy "year end summary" visualisation. And if you want to do a complex filter, can just write an SQL query to the database itself.
- Has a great UI.
- Allows you to write your own import connector and mix 'n' mash between locally-saved CSV/OFX files and APIs, rather than going down the standard route of only supporting Plaid. Current solutions leave users high and dry if the Plaid connector doesn't work well / they want to import old data (Plaid only gives the last 2 years I think).
Re: Ask HN: What are you working on? (May 2025)
#565I am currently working on thethoughtcatcher.com It started as a side project to explore the latest AI trends. Now it’s something we use daily — and others are starting to as well. Thoughtcatcher is a lightweight, AI-powered notes + reminders app that acts like a memory companion. It helps you: - Capture raw thoughts and auto-tag them using AI - Set smart reminders triggered by context and meaning - This was a game ch…
Nice to see an entirely new category of program, based on the core data type "thought", which is what I would call pre-organized information. A spontaneous thought could become a task or a whole project plan, a project or product, once refined, or a memo, even a book. The value proposition is to capture early where other tools force more orgnization before or at capture. So one problem you solve is losing a nucleus of a thought before having time to enter it in a program that requires more metadata. A second way you provide value is that despite being pre-organized, your "thoughts" are already somewhat actionable - this is what using an LLM gives you.
Personal information management is flooded with useless "TO DO" apps that offer little value over plaim text files (which is what I use for tasks), but it appears you have put something together that caters to the informal side of knowledge management in the same sense that Google catered to informal (unstructured) Web pages, overcoming the world of relational databases and forced categorization of library card catalogs.
My recommendation would be to offer APIs and plain text import/export capabilities to grow your app into an ecosystem that can accommodate playing with enterprise tools or homegrown solutions.
I wish you a huge success with this app!
Re: Ask HN: What are you working on? (May 2025)
#566I've been working on a pickleball/badminton (or really any doubles type sport) event manager which integrates an Elo-style statistical rating system (OpenSkill) into the client-side UI. By simply logging the win/loss of every match that happens, it can real-time update and estimate skill ratings in only a few iterations of matches. Its use cases are to manage open play events even with a wide spread of levels, balance the teams of a match, and enforce fair turn-taking.
I implemented a simple first-come-first-serve event RSVP system, and from the event RSVPS it will pull the players into an event management UI from which the organizer can place players into the available courts and arrange matches balanced by level automatically.
The app does ratings calculations fully locally in the browser so it not only works offline but every new match is based on the last rounds' updated player ratings immediately. There is an option to submit matches to be persisted in my hosted service (pkuru.com) so player ratings are retained. The goal isn't to rank players like existing popular systems (DUPR, UTR, etc) but to do matchmaking within an organized session, or before a session.
The web UI and session management app, which is tentatively named FairPlay, is open source on my github. The version I host at Pkuru.com is free to use as well, and currently used by the #1 club (by average skill) in central Tokyo, as well as the highest level club in Thailand.
I don't plan to commercialize it as it is merely a side project but I can't promise the service will always be free to use due to potential operating costs at scale. However, the front-end UI including the session manager is and will be open source.
There's no fundamental reason the system cannot be easily adapted to other sports such as volleyball, soccer/football, or even a 100-person race since the rating system was designed specifically for team based competitions. It can do everything from ranking players in game like PUBG (100 teams of 1-4 players), to 5v4 futsal (asymmetric teams), to classic 1v1 duels as well.
If any clubs or even businesses are interested in using it to manage open play or rec play sessions where people can just drop in and out and be put into a balanced match of the appropriate skill level, please feel free to contact me. The hosted version on pkuru.com is free to use, but I can also assist in my free time on a volunteer basis if you wish to adapt the front-end open source code.
Re: Ask HN: What are you working on? (May 2025)
#567I've been collecting and digitizing vintage print advertisements and publishing them ( https://adretro.com ). I have tens of thousands of ads in the collection and it would take me many lifetimes to complete, but I've been using AI to extract and catalog the meta data. I can get through about 100 ads/day this way. One of my favorite ads, a computer from 1968 that "answers riddles": https://adretro.com/ads/1968-digi-c…
This is awesome-sauce! Any thoughts on an API?
Thank you! Yep I'm thinking about that. Any input as to the use case? The catalog is a full web application I've developed. I'm currently only publishing the best ads to the site for now to not clutter it up. But I could offer an API or access to the full metadata.
Re: Ask HN: What are you working on? (May 2025)
#568SIMD enhanced YAML parser. My original parser didn't have enough MB/s (around 2-70 MB/s), so I'm trying out SIMD. I hope it goes well.
Re: Ask HN: What are you working on? (May 2025)
#569AirPosture: Turn your AirPods into a posture coach on macOS
A macOS app that uses your AirPods' sensors to catch bad posture in real time.
How it works:
Real-time tilt tracking – Your AirPods already have the tech Customizable alerts – Adjust sensitivity so it nudges you only when needed Prevent strain before it starts – Stop neck pain and headaches at the source
Demo:
Re: Ask HN: What are you working on? (May 2025)
#570A platform for project prioritisation. Or at least I thought I was. It turns out that early users really want to use it as a sales tool to demonstrate ROI to their prospects, so I've been pivoting towards that. Check out https://www.reschematic.com if you're interested.