alt Hacker NewsAsk HN: What are you working on? (November 2026)
What are you working on? What have you been curious about lately?
Comments
A suite of high-dopamine games which use smart exercise bikes as the control interface.
When playing Zwift, Rouvy, etc, I find that I get too competitive and ruin longer workouts by trying to keep up with or sprint past people who are much faster than me over long distances.
This is bad for building stamina and endurance, but I get bored, and it's hard to fight boredom and physical strain both at once.
I do not get bored sitting in front of Balatro, Raccoin, Tetris - anything with bright colours and satisfying noises and combos.
The first finished prototype game is very simple: you sort coloured marbles. Higher watt output from the bike means marbles are blown, by wind, upwards as they roll across the screen. Lower watt output means they fall. Steady watt output means they move horizontally. Marble goes into correct hole -> satisfying noise. It rewards steady watt output over time, with minor adjustments, which is what I'm trying to work on right now.
I was pretty amazed by how engaging I found it (total iPad baby brain). The miles really fly by - I get tired before I get bored.
I think that next I'll either do a tower defence game, or a less-dopamine-rewarding game in the style of Kerbal Space Program. Taking your actual weight and watt output, putting a bike with wings into a physics engine, and the gameplay loop being unlocking increasingly large airplane pieces, and designing - then flying - increasingly ridiculous airborne bicycles seems like a compelling idea, though less good for fitness.
The control surface being so limited - essentially one measurement which can either go up or down - is a really fun constraint for creativity.
I always wanted to have all laws to be easy to interpret or at least try to track their changes over time to better understand what’s going on. So I’ve built OpenNorme for the Italian legal corpus here: a single git repo with all laws from 1861 to today with daily scraping from an official api. https://github.com/dej611/italian-laws-history
After every update the github pages is rebuilt with the new data, so you can see some trends in there with some aggs and nice charts. Also, had some fun to build a clippy clone with an in browser SLM.
I’ve been working since the start of the year on a minimalist web reader (https://papercove.app) that I use every day to follow RSS feeds, newsletters and read-later, and that helps keep things from piling up.
I thought I’d get where I wanted much faster, but it turns out I ran into quite a few challenges: it’s quick to build something simple, but it takes much longer to get a reader that’s robust and pleasant to use (extraction, parsing, security, infrastructure, etc.).
On the tech side, I’m using my usual stack, Laravel and Livewire, with some Python for extraction and Go for the email infrastructure. This month I’m working on the beta release, since several people around me were already using it, so I’d like to gather more feedback.
I am working on Raspberry PI OS mod for kids - https://luminum.tech/ I have one user (my kid). The core of the idea is to have one no-limit device at home that is dedicated to learning and not games.
Mother of all VPNs -- https://moav.sh
I used to really enjoy watching Dota 2 tournaments. I also like Bake-Off, and Battle-Bots was a great idea.
I'm tinkering on a project to take two contestants and dump them into a reasonably nicely equipped linux environment (checkpointed fly.io Sprites), giving them a task, and watching them compete in real time to see who gets it done first. All of it streams to twitch.
Here's an example match (I have two agents pretending to be contestants for the sake of the demo: https://www.twitch.tv/videos/2923688013). Basically anything goes; I imagine people will clone+stow their dotfiles, leverage their workflows/harnesses/editor configs, etc. Viewers will get to see how its done.
Next on the todo list is "weight classes" that let contestants compete with a fixed level of harness capability: natty means no agents at all, small local open weight LLM tier only (qwen, gemma, etc), large open weight models only (deepseek, glm, kimi), frontier lite (haiku, luna), frontier (6.1-sol, haiku), frontier plus (astra, fable).
Ultimately I want to rank users in each weight class so people can compete for delivering the most at various cost tiers.
I've recently been making improvements to Larm, https://larm.dev, an uptime monitoring service I've been running for a while now. It's got global probe coverage, automatic/standalone status pages, and lots of alert channels. Pricing is simple and the limits are generous.
It's accumulated a decent number of users now which is exciting.
Under the hood it's an Elixir backend with a network of little go probes running all over the world, all connected by TCP tunnels. All the check result data is stored in Clickhouse. Status pages are generated on demand and served through a CDN.
The whole thing is EU hosted.
AI models have gotten so good, that we felt that what we were missing was the connection to the tools, so I built
No subscription, pay-per-use to allow your AI to access the best tools.
I have developed an app designed to encourage in-person meetups among groups of friends. It makes it easier to organize outings for people who already know each other. You place a pin on a map and share it with your group. Those who already have the app receive a notification. Those who don't can be invited through a universal link. Even without the app they can see the location, the time and the number of confirmed attendees on the web (the app is required to confirm attendance). I am building it on my own with FlutterFlow, Firebase and local SQLite, in five languages.
This month, besides promoting and improving the app, I've had to spend time defending it: bots used the SMS login to trigger 1,744 code requests in nine days. Only three of them ended in a sign-in. I have since added several defenses, including a daily check that compares codes sent and sign-ins per country. None of this fixes the app's core problem. Open the map before any of your friends use the app and you find it almost empty.
https://www.orbitome.com/?utm_source=hn&utm_medium=community...
I was laid off recently, so I needed to update my knowledge to be able to pass interviews. But I’m a slow learner, and I needed to accelerate the process. So I built Tutor, an open-source project that extracts all the concepts (using AI) from study materials (I started with books and online documentation, but recently added YouTube courses as well), creates lessons in simple language, and also tests you. It helped me tremendously to get out of my brain-freeze state.
It uses Luna and Haiku, which have both proven smart enough and are dirt cheap.
Just trying to keep https://www.boot.dev alive and kicking amidst this wave of AI doomerism surrounding... The value of being good at things.
Working on an in-browser update of our k8s course, a new C course, as well as terraform and bash courses coming down the pipe.
Building a development environment to run coding agents in the cloud. There's a lot of other products in this category. What makes ours really special IMO, is the built-in task management system. Your agent can split up large jobs into subtasks itself and run each one on a different agent / model / machine. You can clearly see where things are at right now. In my personal usage, I can run tasks autonomously for weeks at a time this way and just provide steering input to the coordinator.
Would be super happy if you'd check it out: https://podium.do
Over the last year, I've been working on Review Box https://review-box.rstankov.com/ on the side. It is a tool to help me with GitHub Pull Requests.
It has an integrated dashboard for what to review, integrated symbol understanding, a clear split of reviewed/unreviewed files, and a result diff view (which shows me only how files are going to look in the end).
I'm almost not opening Github these days.
I'm building an open-source "spatial workspace for directing AI coding agents like an engineering director".
I got overwhelmed by walls of text of parallel Claude Code sessions. This limited the sessions I could steer in parallel to maybe 5 or 6. I knew there had to be a better way, and I found one: A canvas per project. Work laid out as cards. An agent in a worktree per card. Chat condensed to WhatsApp-style back and forth, not a litany of tool calls. Result is usually a narrated demo video that I can comment on by voice to have the agent go another round. Remote-control from my phone. Remote workers to scale beyond my machine. Everyone I know who tried it has stuck with it so far.
Hand-drawn MMORPG where humans play in one world, bots and AI agents play in another, and they meet in a shared town :) Bots play over plain HTTP, so a few lines of JS or an LLM agent will do.
Trying to make it fun for both normal players and engineers as well.
I’ve been building a little Python tool to extract a MIDI drum track from an mp3 or other song file. Eventually I want to buy an e-drum kit and learn to play along with my favorite songs.
From what I can tell, we’re about even with the current state of the art in academic publications right now, at least on MDB. Runs faster than real time in pytorch on a few years old MacBook.
It even does a decent job creating easy/medium/hard versions from the original “expert” track. I compared ours to some charts from Rhythmverse and while we’re not as good as the official ones, we’re pretty darn close.
Spent some time trying to extract an instrument model for the drum sounds from the recording too, but I think that simply requires too much data for lots of the songs that I like where the mix is really dense with a ton of sound from distorted guitars and cymbals etc. Works a treat on super clean mixes like Back In Black though.
Next step is probably to create my own rhythm game, since my initial experience with Clone Hero and YARG left a lot to be desired.
I've been building AltStack.jp [1], a directory of digital services (cloud hosting, domain registrars, email providers, etc.) that are operated in Japan by Japanese companies.
The site is static, built with Astro. When it came time to host it, I didn't like any of the Japanese options. I could have put it on Cloudflare, but hosting a directory of Japanese alternatives on an American CDN felt a bit off. So I started building my own static hosting service: Kamakiri Pages [2]. It's CLI-first.
Note: AltStack.jp isn't hosted on Kamakiri Pages yet. I plan to switch it over within the next two weeks.
Feedback welcome on both, especially if you know of services that should be listed.
primarily, an agent for real estate agents that lets you manage your CRM with voice/chat, personalized property feeds based on what prospects/clients actually like, automatic routing for setting up pop-by visits or other field work, and showings automations that let you share a showings itinerary, and track both the agent and the client on a map, add coffee stops, and more: https://getcoagent.ai/
an agentic logo generation service that vectorizes into high quality brand assets and gives you a complete package (including favicons, OG images, etc.): https://mklogo.sh/
and a PKMS for macOS with a chat-first interface, allowing you to store any link and chat with your library/knowledge graph, automatically file and categorize it, and more: https://gleanary.app/
I've hated all budgeting apps I've used and found them to be always restrictive, expensive, or envelope (YNAB) focused. I don't really need a "budget" app, more of a personal finance app. I just want to be able to review all my family's transactions and look at cash flow and track net worth, so I built my own. A few of my friends were interested, so I gave them access. I'm seeing more and more of these budget apps bring in LLMs which is such a bad idea. The only AI I built is opt-in and uses Jev AI to auto categorize transactions that are stripped of info.
If any one is curious, it's on https://dumplingmoney.com/
A proportional width font terminal. "i" and "w" shouldn't take up the same amount of horizontal space. We used to have to do that because computers were slow and we counted cycles and every bits mattered. We live in a land of plenty now though so we can spend cycles doing math so that ls -l/top/etc table output is still aligned but "i" and "w" can take up the appropriate amount of horizontal space.
Teaching myself how to build a Spiking Graph Neural Network in Zig, to run on my Brainchip Akida AKD1500 neuromorphic accelerators (and Efinix Titanium Ti60 FPGAs).
Good times, but lots of foundational math I hadn't previously been exposed to.
Some reading material if anyone else is interested in that:
Spiking Graph Convolutional Networks (2022): https://doi.org/10.24963/ijcai.2022/338
A Tutorial on Spectral Clustering (2007): https://doi.org/10.1007/s11222-007-9033-z
Semi-Supervised Clustering with Graph Convolutional Networks (2017): https://arxiv.org/abs/1609.02907
Spectral and Algebraic Graph Theory (2025, draft book): http://cs-www.cs.yale.edu/homes/spielman/sagt/sagt.pdf
I building a service to let different AI agents talk to each other. You can use it to connect your own agents, e.g. Muse and Grok Bot, or make Claude review Codex’s work. Or you can use it to connect to your friend’s or colleague’s agents - think Grok Bot team but without the requirements that everyone has to be part of the same org.
Why not use mail? Well, for one, it’s hard to register many different emails programmatically, and you do want a separate inbox for each of your agents. And you shouldn’t really give your agents access to your primary email: it’s not secure, and you will be drowned in noise past a certain email volume.
There is a lot of competition in this space already; it’s heating up. Wish me luck!
I'm writing a recipe site: https://fantasyrecipes.co.uk/
It's aimed at people who enjoy DnD, and I took inspiration from the Necronomnomnom. All the recipes are narrated by a different character.
I made a website mostly to post coconut water reviews. I never really liked coconut water until I tried an unpasteurized option and that led me to spiral down the rabbit hole of coconut water terroir. As far as my taste buds are concerned, the best coconut water is coming out of young nam hom (fragrant water) coconuts from central Thailand.
I'm not really interested in making any money off of this site at any point but it has been fun to see the Google analytics data as a few dozen people visit it every week to read reviews. I've reviewed like 30 different brands at this point.
I built https://recvmail.net to scratch my itch for a way for agents to sign up for accounts with email verification.
Still working on https://github.com/Disla-Novo/Dimidium_Bellerophon, a DSL for 3D printer automation. I’ve been putting off the networking and printer management side, but I had an idea I’m excited about which is a non-planar slicer that outputs my DSL instead of static G-Code, which could allow more dynamic, point-by-point control. Right now I’m mapping out the features and how it would work. It’s currently an IDE I built on Java (Spark). I’m considering moving the backend to C# but haven’t figured out where to start.
An agent runtime that connects to your existing API or MCP: https://cadenya.com
I felt like all of the plumbing to build agents was insane to get to a hello world state. I just wanted to connect my conversation UI to my API. Hence this product.
I'm exploring using XMPP as a coordination layer for multiple LLM agents.
I'm using Herdr to run multiple agents in parallel, and my goal is to see whether I can get them to collaborate more efficiently.
For example, when one agent's PR gets merged, other agents immediately become aware of it, rebase their own PRs and fix conflicts.
What also often happens is that new tangential requirements and sidequests often arise while working in a session, and I now let agents hand these off to parallel Herdr tabs. The handoff can be done with or without XMPP. These parallel agents announce their statuses in the XMPP MUC (groupchat), which enables them to coordinate.
Technical details:
Converse XMPP client runs as a headless NodeJS daemon with a CLI that can be called by agents.
I created a /coordination skill which is enabled by default via a Claude Code hook. When an agent starts, it ensures the Converse daemon is running and then joins a password-restricted and members-only MUC where all the other agents in that Herdr workspace are as well (and also myself). I can then assign that agent tasks via messages in XMPP (DMs or groupchat) or from Claude Code.
Once the agent has the task, it announces it in the MUC so that other agents are made aware. When sidequests or requirements arise in that session, it can ask in the MUC whether other agents are idle and able to take over that parallel task.
I've been building the Converse daemon and CLI and have now been experimenting with the workflow and working out various kinks and use-cases.
There are various other XMPP features that can be explored further. I'm thinking of using PubSub with a web UI with vertical swimlanes which can show me which agents are currently active, what they're working on and for how long.
I'm using Claude Code because that's what we use at $WORK, but this should be independent of any harness and part of the point is create neutral tools not locked into any particular inference provider.
Trying to make my favourite tactical FPS game - Rainbow Six Rogue Spear in Java with some modernisations.
A vim-inspired terminal PDF reader using MuPDF and Kitty Graphics Protocol, mostly for my personal use since I want to use vim bindings for reading PDFs.
Lately, I've been redoing the entire writing screen in Gorby [1], trying to make it more useful and more pleasant to use. I've added better themes and brought some features and text checks from other parts of the app into the writing screen as well. I'm also working on integrating Harper.js for grammar checks. So far, I've only included it in the desktop app (yeah, it's built with Electron…), and it seems pretty decent. The desktop app itself is new as well. Generally, I'm trying to make the app useful for writing, not just for finding possible improvements in the text or getting its readability score.
I also made some improvements to my support workflow app, Support Toolbox [2], but haven't released those yet (hopefully this week). It now also includes text expansion, which I was really missing from my workflow and had been using another app for previously. At this point, it does pretty much everything I need it to do. There are some smaller improvements I need to make, but even if no one else ever starts using it, it helps me tremendously at my day job!
[1] - https://gorby.app/
I've been burning the midnight oil and plenty of tokens on two projects.
Project #1: The first project is MiSTerVision, a CRT-friendly Jellyfin and Plex client for MiSTer FPGA built in Go. I have a MiSTer FPGA connected to a late model Sony Trinitron CRT and wanted to watch movies and shows I've ripped to my Plex server.
I originally contributed PRs to the MiSTerFin project, a Jellyfin client for MiSTer, but the project was stagnant so I forked it, ported it to Go, and was off to the races. I added Plex support, improved the UX, improved performance, filled in a bunch of compatibility gaps in the Jellyfin support, and released it for use by anyone with a MiSTer.
Watching old-school cartoons, like Dungeons and Dragons (my favorite) on a 4x3 CRT in better fidelity than I ever had as a kid is sublime. It's had 25 or so public releases over the last few months.
https://github.com/trentnix/mistervision
Project #2: The second project is HyperServer. HyperServer is a Go framework for server-rendered apps with HTMX that I've been working on off and on for a few years. I was originally inspired by Pagoda and really like the architecture and ambition of Pagoda but didn't like its dependence on third-party libraries.
HyperServer uses standard net/http handlers and html/template, requires no ORM, and makes authentication, sessions, and storage optional. It’s intended as a reusable library rather than a starter repository you copy.
I haven't been confident enough to make a "release" for it yet, so it's labeled as "experimental". But I have built a functional (and fast) blog on it. I plan to use it as a platform for a couple of other web projects I'm scheming. Hopefully that will prove its utility and I'll make an "official" release.
That said, anyone interested can peruse it and use it as-is. Any feedback would be welcome.
I'm doing something ambitious. I'm taking a radical approach to improving rust compile times. The reason it is radical is it involves building a brand new rust compiler. This was inspired by a project I started[1] about six months ago to teach agents to build C++ compilers from scratch. I wrote a (long) series of blogs about the process[2].
After iterating on many agent built C++ compilers, I now have a version that compiles about 4x faster than GCC and Clang[3]. I think i can extend the same approach to rust. I'm still a ways from the rust self-hosting stage but early results are promising.
[1]: https://github.com/vishvananda/cppgm-extended
[2]: https://medium.com/@vishvananda/i-spent-2-billion-tokens-wri...
[3]: https://medium.com/@vishvananda/how-i-built-a-compiler-that-...
I'm working on an integrated system with Home Assistant for making it widely available and more affordable for everyone. The solution consists in doing a touch panel for every room that can run predefined actions. We've tried the mobile app / web app only but I think this can fit better. I'm trying to do a business out of it. I have the alpha version that works and now I'm working on redesigning the panel.
I also consider doing own devices in the future. Providing an affordable solution for everyone.
Project is Rust + Slint.
If someone is interested in changing the smart home scene my email is in my profile!
I've been working on Semiotique (still very early stage).
Right now it's a python library, but the vision is to build a new framework / SDK / programming language which allows you to seamlessly mesh "plain old code" with LLM evaluations, while having a runtime which does "smart" context engineering in the background in order to steer the LLM and improve its correctness.
The end-goal is for this paradigm to enable new kinds of agentic applications, which leverage the flexibility of LLMs while retaining the determinism of regular code.
My partner and I are both software devs. She quit her job a year ago, and has been working on a few micro-businesses to try a) learn some new skills and b) diversify our sources of income. So far this year she has built:
* https://unofax.com - A simple a la carte fax service
* https://unopost.com.au - A way to send physical mail online (and avoid standing in line at the post office)
* https://snailgram.com.au - Custom greeting cards, with the message inside written by a 3D printer holding a pen.
There are a handful of others too, but these are the ones that are live so far. I've contributed between 5-40% to these projects.
I'm working on dushed[1]. It's a personal knowledge management web app that combines flashcards, bookmarks, notes, web feeds, files and Kindle highlights into one.
Over the years, I've used spaced repetition to learn things and to remember them. But the material for my flashcards came from everywhere: books on Google Drive, blog posts in a bookmark manager, Kindle highlights, Wikipedia, and notes spread across email and notebooks. The cards were in one app and their sources were in five others. This was overwhelming so I'm building dushed. You can access all your knowledge sources(bookmarks, feeds, Kindle highlights, files, flashcards, etc) in one place and also link sources to one another.
Me and my friend are building a platform for agricultural / water treatment / fertigation machinery.
Traditionally these kind of machines would use a PLC, and then get connected to Scada and Internet, but that's expensive.
We have built Linux based controller that can extend into any number of AO, AI, DI, DO inputs and control the machine. It's very cost effective, and unlocks a lower bottom of the market that before didn't even dream about installing connectivity inside each machine.
Our first project is https://waboost.com/
It's a water treatment machine, and we are looking for more clients.
Right now we are talking to grain separator company, and one fertigation company.
Stack: Go for Cloud and machine backend React for Cloud and Machine UI
Pretty minimalistic
As a heavy LLM user in my daily work, I've been working on optimizing agent memory on complex projects. That resulted in this simple Rust-based tool called Wayfinder: https://github.com/zaterka/wayfinder
I've been trying to improve agent knowledge over a codebase, and existing approaches weren't giving the result I wanted, and I wanted to keep it simple. In the end, I added a semantic layer on top of an LLM Wiki and it saved 30% of token usage in multiple benchmarks I ran. It also adds some nice features to deal with concurring agents updating the shared memory, to ensure all information there is reliable and fresh.
It's published to PyPi, so an "uv tool install wayfinder-memory" already does all the work, and it's compatible with almost all Harnesses.
https://github.com/mquan/netflix_challenge
I'm revisiting the Netflix Prize challenge with LLM. So far I've hit the 9% improvement mark within a couple weeks. I seem to have hit a wall in the last week or so with no further progress. I'm hopeful to hit the 10% goal with one of the many experimentations though.
I am working on Liitmik, www.liitmik.com. It is a win/mac/linux desktop app that lawyers can point to their filesystem-based project catalog and it will then enable them to spin up AI agent sessions quickly by selecting the right project from their catalog via autocomplete search and starting the agent (Claude Code, Codex, OpenCode) in an isolated VM. So it provides lawyers strict context management and strict agent isolation on their own machine in a way that makes creating sessions fast. I firmly believe that lawyers (and other similar professionals working with sensitive data) do not need custom agents, they just need a way to use the widely used models and harnesses in an environment that is more strict than the defaults provide.
I have 10 years of experience as an attorney doing mostly dispute resolution. And about 15 years in software development. Liitmik is the product of these two.
Building https://birding.live/ a birding site, that helps people look up bird sightings in their area, specifically those classed as rare. Connects to a whole lot of open data sources (eBird, OSM, Nominatim etc.)
Will remain free, planning to add some kindof mailing list for alerts when birds are spotted in areas of interest.
Been working on https://backlog.fyi, a personal task manager, for the last few months. I tried quite a few approaches first. The simple ones, like Microsoft To Do, Google Tasks, text files, or even good old Post-it notes on the wall, were too basic, and it was too easy to just forget about them forever. The sophisticated ones were overkill, asked for too many permissions, and their notifications quickly turned into noise. So I made one that fits me. It’s focused on personal task management, it has a weekly planning session so I don’t forget things, and adding a task is as easy as forwarding an email or sending it on Telegram. It’s also privacy-respecting and fully open source (https://github.com/Itaypk/Taskboard) if you feel like self-hosting it. It's a fun project, not for profit - no ads/trackers/subscriptions page. It’s still a work in progress, but I really like it for my day-to-day.
Built with Kotlin/Spring Boot/React/PostgreSQL, because boring works well.
- https://probl.dev: A "probabilistic" programming language where we can do things like `if 45% { x } else { y }`. It works by creating many weighted worlds, that are then merged when identical. Operators like `observe` can remove possibilities. There is also a sampling mode that instead run many simulations to get an approximate result.
- https://github.com/garlab/gloss | `brew install garlab/tap/gloss`: render markdown, sources, images, chess and go (game) records in the terminal. It's built for libghostty but fallback on other terms.
aix - Like the npm CLI and package.json, but for AI config. Allows standardizing your AI config to share with others, and defining it all in one spot but installing to Claude, Codex, Cursor, etc.: https://aix.a1st.dev/
Classifieds.fyi (https://classifieds.fyi) - A way to create private, invite-only classifieds for your community. Supports buying/selling, job boards, and vacation rentals, with support for housing coming soon.
A Tauri 2 CLI / MCP that allows your agent to debug, take screenshots, run JS, etc. inside a Tauri app: https://hypothesi.github.io/mcp-server-tauri/
Safe Drive - Mobile app for my car. I learned driving couple of years back only. Everyday, I travel same route. So decided to create a small mobile app first, which will provide me feedback about my driving. eg. which part of the route im doing hard brake, which area im driving smoothly. etc. app will warn me few seconds in advance like "Warning: Often you perform hard brake in this region" etc. Still in early stage. I'm software guy, Im exploring still hardware parts, which could help to gain confidence with either side of car, (how far the objects etc). I know high-end cars may come with these by default (360deg view etc). Later if possible I plan use something like openauto/raspi if my interest persist still then.
I'm working on getting some old flip dot displays up and running again, more precise the single 5 x 7 modules that bigger displays are made of.
There are lots of these projects around the internet but I found them to be hard for an electronics noob like me, and so I put the modules back in one of the "will fix later" drawers. With the current state of AI I decided to give it a go.
My Fluke and me did some measurements and Claude did the hard work afterwards. Prototype worked on a breadboard so I let it cook further and now I've got some PCBs coming in from China.
Still got a lot of respect for soldering some (quite large) SMD components but I think I'll get it done.
My plan is to revive as many modules as possible (got around 30) and build clocks from them to gift away to family and friends.
What are you working on? https://wpmaven.ai Describe your business and WPMaven.ai creates it, runs it and grows it for you. It does all your marketing, including UGC videos on all social media, sales follow-up and closes sales for you.
You can even manage your business on your WhatsApp! You receive your weekly report in your phone directly and you approve or edit anything that goes online (because yes, AI is not perfect yet).
We launched recently and are polishing everything while finding the optimal verticals (if you guys know WHERE would you use it, let me know!).
I’ve been working on BilingualBulletin, an iOS app for learning a language from real news, books and podcasts. The coolest thing about the app is that nearly everything runs on the phone, and there’s no account. I’m using Apple’s on-device translation, you can ask questions to Apple’s foundation models, and audio and video is even transcribed locally.
It supports 19 languages, and comes with a lot of built-in RSS news and podcasts feeds, though power users can add their own. You can play audio and follow along word by word.
I’m looking for more beta testers at the moment. You can sign up at https://BilingualBulletin.com. It requires iOS 26.
I’m curious if anyone else has been building on the Apple Foundation Models or other onboard AI frameworks? They’re surprisingly powerful, but with weird limitations sometimes.
https://minihomelabs.com/hwdb - Basically the PCPartPicker of used OEM hardware.
It's harder than it should be to compare old pcs for home labbing so I've been building this to scratch my own itch. Right now it's more focused on specs but I'm working on expanding out to to include things like viable upgrades and explaining the different model lineups.