alt Hacker NewsAsk HN: What Are You Working On? (June 2025)
What are you working on? Any new ideas which you're thinking about?
Comments
I'm working on a new kind of encyclopedia and reference website that is way more engaging and rich than a single page of text on every subject like Wikipedia. Back in my childhoold we had Microsoft Encarta on CDROM and it was a very compelling multimedia experience and I want to emulate something like that on the modern web by combining video, sound, images and text in a more compelling user experience with great discoverability. I've been working on the first version for about eight months and I hope to launch it next week at https://reference.org
Mostly writing for myself; I should really convert some drafts into proper blog posts because I'm really interested in discussing my ideas with others.
I've been thinking a lot about the current field of AI research and wondering if we're asking the right questions? I've watched some videos from Yann LeCun where he highlights some of the key limitations of current approaches, but I haven't seen anyone discussing or specifying all major key pieces that are believed to be currently missing. In general I feel like there's tons of events and presentations about AI-related topics but the questions are disappointingly shallow / entry-level. So you have all these major key figures repeating the same basic talking points over and over to different audiences. Where is the deeper content? Are all the interesting conversations just happening behind closed doors inside of companies and research centers?
Recently I was watching a presentation from John Carmack where he talks about what Keen is up to, but I was a bit frustrated with where he finished. One of the key insights he mentions is that we need to be training models in real-time environments that operate independently from the agent, and the agent needs to be able to adapt. It seems like some of the work that he's doing is operating at too low of an abstraction level or that it's missing some key component for the model to reflect on what it's doing, but then there's no exploration of what that thing might be. Although maybe a presentation is the wrong place for this kind of question.
I keep thinking that we're formulating a lot of incoherent questions or failing to clearly state what key questions we are looking to answer, across multiple domains and socially.
I'm building a web game, inspired by GeoGuessr, but for robotics: find the robot's axis configuration to reach a specific target.
It's still a WIP... lots missing, but happy to get any feedback!
The stuff I'm working on never feels worth sharing, but I am doing a lot of computer stuff lately. It's kind of the year of moving towards declarative setups for me.
- Migrating to Niri on my laptop and re-evaluating my literate config approach, switching from xkb configs to kanata and a few other QOL changes to make my tooling more composable and expressive
- Shoring up my blog / media sharing infrastructure (migrated to a landing page on an s3 bucket, with different prefixes for several different hugo deployments for different purposes, still need to get better about actually posting content)
- Preparing to migrate a bunch of my self-hosted services to a k8s cluster which can can be fully deployed locally for testing and defined in code. All this is managed through argo and testable with localstack and crossplane for some non-local resources
- Attempting (somewhat unsuccessfully) to setup a nixos config for a bunch of services that just don't feel right to run in containerized stack that I want to live in ec2 and have as close to 100% uptime as possible (uptime kuma, soju/weechat relay/bitlbee, conduit, radicale, agate, whatever else I think of that is small and has a built-in nixOS service module. Thinking about some kind of RSS aggregating solution here as well)
- Experimenting with vibecoding by trying to get an LLM to do the legwork to build a TUI interface to ynab using rust (which I don't know how to write)
I'm hoping that by the end of this summer most of the tooling I use for most things will be way more concrete and seamless. I also want to get my workflows down and get on top of converting at least a few the ~100 draft blog posts I have laying around into something I can actually post. Ditto for my photography albums, which are not yet organized into coherent groupings or exported for web.
Working on tail calls for TXR Lisp. Current release provides self tail calls only; and certain cases don't work, like applying in tail position. Plus there is a shadowing bug. These issues are addressed already.
Tail calls between different VM functions are the next challenge. I'm going to somehow have it allocate the VM instance in the same space (if the frame size of the target is larger than the source, "alloca" the difference). The arguments have to be smuggled somehow while we are reinitializing the frame in-place.
I might have a prefix instruction called tail which immediately precedes a call, apply, gcall or gapply. The vm dispatch loop will terminate when it encounters tail similarly to the end instructions. The caller will notice that a tail instruction had been executed, and then precipitate into the tail call logic which will interpret the prefixed instruction in a special way. The calling instruction has to pull out the argument values from whatever registers it refers to. They have to survive the in-place execution somehow.
I'm working on MedAngle, the world's first Super App for current + future doctors. An invite only platform, which has everything people in medschool/dental school and recent graduates need. From analytics, quizzes, summaries, x-rays , videos to tens of thousands of questions, ~100k+ students/doctors have solved over 100m questions, spent 10s of billions of seconds, and growing!
We're also working on the Premed Super App, same thing for people taking medical school entry exams like the MCAT or MDCAT.
I get to work with a bunch of top notch students and doctors, and I myself am the first ever full-stack technologist who also is a doctor in Pakistan, a country of 250 million people.
I'm working on an app that I use to track my sets/reps. I've added in a trends screen (both for muscle groups and for individual exercises) to check on if I'm actually progressively overloading on a consistent basis. I'm currently adding an AI wrapper around the data to figure out better insights about certain exercises, what I'm missing, etc.
Working on Cursor for Excel: https://www.tryalphaexcel.com/
As there is no open source version of Excel except Libreoffice, working to build the core Excel functionality with other open source packages. Then bringing in agentic editing functionality for real world data.
What is also has been interesting is to introduce banker/consultant formatting guidelines to the agent and making it beautify its work whether in tables or models.
Hej, I made FisherLoop[1] to learn Swedish. FisherLoop are interactive audiobooks where I use TTS with word level timestamps to highlight the words as they are spoken. This helps me pick up on pronounciation and grammar in a, for me, natural way. Additionally, I added flashcards from the books + word lookup. I am adding new books right now. If you have any requests: public domain books, which are around one hour reading time let me know :)
I am using cerebras for book translations and verb extraction and all LLM related tasks. For TTS I am using cartesia. I have played around with Elevenlabs and they have slightly natural sounding TTS but their pricing is too steep for this project. Books would cost a couple of hundred euros to process.
Syntax highlighting caught my interest after I created a data format. I stumbled upon TextMate grammar bundles which are supported in some editors and created a bundle that works with three of them. The gnome text editor uses a different language definition format for which I created a grammar file as well. [1]
To highlight the syntax in the browser I checked out the CodeMirror project that uses Lezer grammars. It is very flexible and allowed me to implement additional features like custom folding. [2]
I would also like to create a grammar for tree-sitter, finish the Java implementation and documentation of the ESON parser before I try to implement it in other languages.
[1] https://gitlab.com/marc-bruenisholz/eson-textmate-bundle [2] https://gitlab.com/marc-bruenisholz/eson-lezer-grammar
I'm building https://prijm.com which is a minimalist link sharing and post creation platform with custom feed and notification support for your activities. Here are some of the features:
- Supports markdown every where, even in your comments and replies.
- Get notified.
- Personalized feeds.
- Lightning fast & mobile first.
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).
Currently building a SaaS for product management to make up for some of my own defects. Looking for beta testers if anyone is interested in trying it and providing feedback. There will definitely be bugs but I plan to start using it for myself to help me keep up on some projects I'm working on. The big thing I'm excited about is AI Risk Analysis, which will review the dependencies and hopefully catch bottlenecks and issues before they cause misses. Current URL is below. Name is pending. Just had Google tell me something that sounded cool.
Still working on Orbiter[1] but with some big changes.
Orbiter started as static website and web app hosting, but we recently added what we call server functions (what you call serverless functions). So now you can build full stack apps. Our preferred development stack to support is the new bhvr[2] stack.
I've been working on implementing @mpweiher "Storage Combinators" [0] and "polymorphic Identifiers" [1] in Eiffel [3].
Currently I'm stuck implementing a storage combinator with EiffelWebFramework[4]
[0] https://dl.acm.org/doi/abs/10.1145/3359591.3359729
[1] https://scholar.google.com/citations?view_op=view_citation&h...
[2] https://en.wikipedia.org/wiki/Eiffel_(programming_language)
A job feed for remote jobs - https://tangerinefeed.net/
This is something I’ve needed myself over the last few years as jobs become shorter and shorter lived. Keep on improving it as some kind of compulsion.
I’ve been working on an app called Lång. It’s a calm daily spending guide – shows you what’s okay to spend today, based on how much needs to last how long.
The idea came from noticing how most people manage money day to day: checking their balance, adjusting by feel, trying not to drift. There are tons of tools for planning or categorising, but not much that fits that kind of improvised pacing.
Still early, but trying to shape it around those habits – to make something simple and steady, that supports how people already do things.
Just vibe build a Table Viewer based on Markdown table syntax you can share via links. https://v0-markdown-to-table-view-snipped.vercel.app/ Eventually I would like to make it embeddable into LinkedIn posts.
I'm working on an MCP to give your coding agent the ability to generate on-demand Mermaid diagrams about anything in your codebase. Among other benefits, it is very helpful for spotting unnecessary code or architecture that can accumulate while vibe coding.
https://www.npmjs.com/package/@mindpilot/mcp
Claude Code Quickstart:
``` claude mcp add mindpilot -- npx @mindpilot/mcp ```
I'm working on FeedSync, which polls RSS feeds and pushes new entries to Discord.
It's built on Phoenix/Elixir and has been pretty rock solid.
Sign up currently requires a card number, but that's something I'm hoping to fix in the next week or so. If you do happen to sign up, I'll set your account to an unlimited free trial ;)
I'm building an Electron app that integrates with Shopify and QuickBooks Online to streamline operations for artist-led product businesses. As a jewelry maker, I'm automating the spreadsheets and photo archives I rely on to run my studio. I want a single source of truth for all the data scattered locally and across platforms: tracking material prices, calculating costs and labor, analyzing marketing ROI, and accurately updating cost of goods sold when a piece sells.
I'm tinkering with relative positional encoding by trying to integrate acoustic features directly into it.
More specifically, I'm trying to use pitch (F0) to dynamically adjust the theta parameter in rotary positional embeddings, so the frequency of the positional encoding reflects the underlying pitch contour of the speech and instead of using a fixed unit circle (radius=1.0) for complex rotations, I'm trying to work out how to use variable radii derived from the pitch. The idea is to create acoustically-weighted positional encodings, where the position reflects the acoustic salience in the original audio. https://github.com/sine2pi/asr_model
I'm still working on hcker.news, which first started as a more configurable hacker news frontpage, but has turned into a thing that I've found to be quite helpful at content discovery.
I recently by request[0] added a cohesive timeline view for hn's /bestcomments. The comments are grouped by story and presented in the order that they were added to the /bestcomments page. It's a great way to see popular comments on active topics. I'm going to add other frills like sorting and filtering, but this seems to be as good a time as any to get some of your thoughts!
You can check it out here: https://hcker.news/?view=bestcomments
[0] https://news.ycombinator.com/item?id=44076987 (thx adrianwaj)
Hey HN – I'm working on OneBliq https://onebliq.com, a lightweight tool to help teams plan and track Azure costs collaboratively, without the usual enterprise overhead.
We built it because managing cloud budgets often turns into a spreadsheet mess, or worse, a never-ending consulting engagement. OneBliq lets you:
* Split and allocate Azure costs by cost centers, teams, or projects
* Visualize current spend and attention areas at a glance
* Experiment with plans and projections without complex tooling
* Skip sales calls and long onboarding – just install and kick the tires
It's still early, but we're seeing traction with teams who want clarity without complexity. Happy to answer questions, share more, or get feedback.
Would love your thoughts – what would make a tool like this useful (or useless) for you?
Sometimes you wish you had best of hn summarised somewhere but that 'best' may be different to different people based on their interests. I'm therefore collecting links I find important, from hn and beyond.
Thinking about:
How will various Human Computer Interaction change as many of current apps (which are screen based UI with some background code) simply get replaced with chat/voice/gesture based requests to LLM
I've been digging into WebRTC to understand how it works under the hood. My goal is to eventually build a lightweight media server with SFU capabilities — but focusing on the protocol level, not just using libraries.
To get there, I'm breaking it down into smaller projects. The first one: a basic WebSocket signaling server written in Rust (based on RFC 6455). I'm also learning Rust, so this was a good way to build something real while figuring things out.
On the frontend, I used Angular and just the RTCPeerConnection API — no external WebRTC libs. The focus for now is just signaling: how peers connect, exchange offers/answers, and so on. No media or security handling yet — that's the next step.
Here’s a short demo video on YouTube https://www.youtube.com/watch?v=V_qdW2JchbU It’s not production-ready yet. Right now, each WebSocket connection spins up a new thread, but I plan to rework that using something like Tokio for better performance.
Curious to hear thoughts or suggestions.
My team and I are building a web app that enables any business to chat with any other business in any language. Details:
It's B2B only - can't register with a free email provider, gotta own a real domain -Therefore identities are collective - companies, not company employees --Therefore all interactions are persisted at the org level rather than assigned to individual inboxes
-It allows you not just to talk but also to work together on contracts. We built a contract parser that turns contract clauses into smart, plain language objects
We're calling it Geneva and doing a friends/family/acquaintances exploratory release as I type this.
http://genevabm.com http://x.com/genevab2b https://www.linkedin.com/company/genevabm/
I'm working on a free high performance address matching (geocoding) library. As it turns out I blogged about it just today: https://www.robinlinacre.com/address_matching/
Having worked on various products and startups before, I want to make it easier for solo-founders and small teams to understand website traffic, conversions, product usage, errors their users are having and provide support to their users, without having to intergrate and maintain multiple disjointed tools.
That's why I am building Overcentric - a simple and affordable toolkit that combines web & product analytics, session replays, error reporting, chat support and help center - all in one place.
Been building it and testing with several startups and improving based on their feedback. I am also using Overcentric for Overcentric itself, so I always get ideas for improvement.
What's next: more tools that are useful for startups are on the roadmap and I am exploring how LLMs can be further utilised (apart from support, session replay summaries, aiding in writing help center articles) and refining pricing.
Check it out at https://overcentric.com/
Would love to connect with other SaaS founders and have Overcentric help them grow their startups.
I’m working on a desktop app called With Audio https://desktop.with.audio a one time payment desktop app.
— it turns ebooks, articles, and documents into synchronized audio with real-time text highlighting. It’s great for people who prefer listening while reading (or want to stay focused), and it works fully offline with a one-time purchase — no subscriptions.
I’m bootstrapping it and trying to figure out how to market it effectively. So far, I’ve had some traction and early sales just by posting on Reddit, but I’m still learning the marketing side — especially how to reach people who’d benefit from it most.
Would love to hear how others approached early growth for similar bootstrapped tools.
I'm building a budget app for my wife and myself.
Basic goals:
- Web based for zero update latency
- Have it work offline
- Automatically import transactions from my banks
- No running/hosting cost
- Secure
Tools used so far:
- InstantDB for the datastore, providing the offline capability too
- A gmail account that automatically gets forwarded bank alerts for purchases
- Gitlab.com w/scehduled pipelines for cron based email-syncing
- Netlify for the free hosting
- InstantDB magic codes / email links for securing the data
I'm at the point where I can track and categorize purchases, including split transactions.
Next steps:
- Add in date ranges for reporting / data views; e.g. show expenses incurred in a one month period instead of for all time.
- Add in planned / project transactions for month forecasting
- Statement import & import reconciliation and statement reconciliation
- Scrape company specific digital reciept emails (like Amazon) to autopopulate more transaction data
And that'll be the end of the stuff I can do for free. I think I will add features that require money and/or dedicated hardware though:
- OCRing receipts -> autopopulated transaction data / description
- Using chatgpt to suggest categorizations
- Scrape extra data from my bank sites, like physical addresses of entities involved in charges.
Writing multisynq.io apps in newswpeaklanguage.org. Two game changing technologies now working together. My dream Frankenstein has finally come to life! I (re)implemented my toy application, The Chalculator, to yet another platform (the 7th or 8th maybe?) [1] I added comments to my blog _without any backend whatsoever_ [2] I built a counter app using Newspeak's Hopscotch GUI components to respond to Multisynq (javascipt) subscriptions (the final milestone) [3] 1) https://youtu.be/MPW5L1gj4Dw?si=Lms3odwQElZu_mLu 2) https://youtu.be/8gcCIaXmQjU?si=Qg2XSDu6mIhSAd9d 3) https://youtu.be/2-tZzB8MAX0?si=6nZ1nD6d4isdQd-5
I’m working on https://finbodhi.com — a double-entry personal finance tool where you own your data. It’s local-first, syncs across devices, and everything’s encrypted in transit.
It helps you track, understand, and plan your personal finances — with a proper accounting foundation.
It's interesting in many way. Using double-entry (it's a perspective shift), the technical challenges of building local-first app, UI/UX & visualizations, privacy and more.
I added basic podcast support to my rss reader and now I am working on an android app. https://ivyreader.com
Not a technical project in the typical sense, but I finally started working on a satirical faux-exposé series about my years working in the industry.
If anyone's interested in that kind of thing:
https://massiveimpassivity.substack.com/p/softcore-how-nobod...
I'm working on a simple, local storage budgeting app called "Wasa Budget". I wrote it because I got tired of tracking my budget on excel sheets. It's written in flutter, it works well enough that I was able to entirely ditch the excel sheets now.
I want to publish it on Google play, but I need testers. If anyone cares about budgeting, I'd love to get some feedback.
Here's the app link: https://play.google.com/apps/testing/dev.selfreliant.wasa_bu...
I don't think you can download it without being added to my testers list though. Send me your Gmail address if you're interested!
Continuing to plug away at Trilogy, a better SQL for data consumption and analytics. Getting closer to core feature completeness, at which point can pivot to focus on integrated visualizations + pre-processing/ETL.
Most recently have been focused on better geographic visualizations in the public studio for people to experiment - getting decent automatic lat/long, want to have easy path visualizations (start/end, etc). More AI-accelerated options as well, especially around model authoring.
Repo: https://github.com/trilogy-data/pytrilogy Studio: https://trilogydata.dev/studio-core/
Newly Registered Domain block lists for PiHole[0]. This has worked hands-off for the last two days after a couple of weeks of tweaking. My PiHole blocklist is currently a bit over 14 million domains, of which around 8 million are NRDs.
My next item is to add AbuseIPDB IP addresses to my "Uninvited Activity"[1] IP address blocking system, implementing xRuffKez's script here: https://github.com/xRuffKez/AbuseIPDB-to-Blackhole
Unfortunately, but also understandably, AbuseIPDB limit their free-access (account required) API to 10,000 IP address records. So I might be putting it into a database to hopefully aggregate multiples of the 10k results if they're not always the same 10k.
I've started to make EDM since 2 weeks ago. I've always wanted to do it and now I finally get to spend some time to do it. I'm having tons of fun.
https://thecoderssage.com Trying to build a platform for creating specific tailored coding lessons from prompts or your own code, with interactive coding, quizzes and exercises.Kind of youtube shorts but for coding lessons, on demand.
Working on https://www.space4ads.com - an Out of Home Advertising Marketplace that connects advertisers with underutilized commercial spaces. Think Airbnb, but for ad placements. We're targeting vacant storefronts, construction walls, cargo vans, box trucks, 18-wheelers, and basically any commercial space with foot traffic or vehicle visibility that's not being monetized for advertising.
Been working on a small team exploring what the intersection of coaching, employee engagement, and AI looks like. (https://engage.myemmaai.com/)
Most employee engagement software is just placation for HR. When it's common that the lowest scoring question on feedback cycles is "I believe that action will be taken based on the results of this feedback," there's something fundamentally broken with how companies handle feedback, and how the tools their given enable them to react to it.
Our end goal is to help leaders and managers identify problems with trust and communication within a team. The reality is, 90% of the time, the problem lies with the leadership itself. We're trying to provide both the tools to diagnose what the problems are, and frameworks for managers to fix them.
I just started on an old-school forum software. Go + Sqlite. Good old server-side HTML templating.
Why? I don’t like Discourse and Flarum that much. I want an even simpler solution with fewer bells and whistles.
But I guess the market is dead anyways for forums. I might replace my phpBB instance that has been running for 15 years.
I'm building the Enact Protocol: https://enactprotocol.com
Turn any command into an AI-discoverable MCP tool with a few lines of YAML:
name: hello-world
description: "Greets the world"
command: "echo 'Hello, World"
A free meeting-scheduling tool with all the pro-sounding features without any restrictions or commissions. Think Calendly, but free and much better.
Writing the last chapter of my book myindiebook.com for the official release in July.
It's the story of how I quit my job ($100k/year) with no experience, foundation or advantage to build a profitable one-man business 3 years later.
I'm putting the finishing touches on my free daily word game, Omiword[1][2]. I had it basically finished, with the option for players to make a one-time $5 payment to unlock access to the archives, but then Stripe shut down my account, claiming it was a "restricted business".[3] I'm now reworking it to try to fund it through Patreon, we'll see how that goes.
Side project to enable speech-to-text functionality on my Ubuntu laptop so I can prompt AI coding tools fast:
https://vibed.pub/ "vibe blogging" platform, think bolt.new x substack.
Waitlist is open, /blog contains example posts
https://www.dreamsign.ai/
An AI-native DocuSign
It's been around a month I've been working on it. Struggling with getting people to actually use it - this week I've set the ambitious goal of 10 new contracts sent *and completed* by people I don't know (last week's was 10...by people I do know).
It's hard because I feel I'm in a weird hole - in order to have a good product I need people to use it and give me feedback, but in order for people to use it and give me feedback I need a good product. It's like wth!
Another thing I'm struggling with - enjoying the process. I get daydreams like mad. I feel I'm always living in the future in some way, especially with this software, and it's taking away from being present in this work. Which sucks, because I want to be excited to *work* on this and NOT fake my own excitement towards this as a manifestation of my greed to get rich off it.
But MAN am I greedy. It's ugly sometimes, to myself.
But god how I love to work on software also. How I love making stupid bash commands on my terminal. How I love to feel like the old gods, who conquered the infant digital world.