Live data from Hacker News

Ask HN: What Are You Working On? (March 2026)

news.ycombinator.com

951–960 of 1001 posts

Re: Ask HN: What Are You Working On? (March 2026)

#952
Building motivational-quotes.net — an AI-powered quotes content site designed to scale.

The core pipeline: AI generates structured quote pages → automated image synthesis with custom JSON templates (typography, gradients, layout) → Pinterest distribution for social traffic alongside Google SEO.

Currently in English traffic validation phase. Interesting challenge has been tuning the image generation — switching from monospace to serif fonts made a surprisingly big visual difference. Also figuring out the right balance between AI-generated content volume and quality signals Google actually cares about.

Next steps: Spanish/Portuguese expansion, then launching inspirationalquotes.app as a paid AI quote image tool for social media managers and HR teams.

Happy to discuss AI content site architecture or Pinterest SEO if anyone's been down this road. https://motivational-quotes.net

Re: Ask HN: What Are You Working On? (March 2026)

#953
Building a macOS app launcher called AppTela (https://apptela.com). Apple removed Launchpad in Tahoe and the replacement is pretty bare, so I built something with more depth and honestly better than most alternatives out there (IMHO). You can organise apps across screens, categories, groups and stacks. The stacks concept is the bit I'm most pleased with - pile apps on top of each other visually, or create launch stacks that open a whole set of apps at once. Also added a shadow engine where icons cast shadows based on the time of day, which was a fun experiment.

Just launched on the Mac App Store last week, now figuring out the discovery side which is the hard part. Turns out that doing the best App doesn't mean much if no one knows it exists!

Re: Ask HN: What Are You Working On? (March 2026)

#954

  I am working on Plain Invoice (https://plaininvoice.com) — A free invoice generator for freelancers and small businesses. No signup required — just fill in your details, pick a currency (40+ supported), and download
  the PDF or email it directly.

  I built it because every invoicing tool I tried either required an account, had a paywall after 3 invoices, or was way overbuilt for someone who just needs to send a clean invoice. This keeps it simple
  — pick your invoice type (hourly, product, or fixed price), add tax/discounts if needed, and you're done.

  Would love any feedback from fellow freelancers or indie devs.

Re: Ask HN: What Are You Working On? (March 2026)

#956
Working on Engram, a cognitive memory system for AI agents. Instead of vector DB + semantic search, it uses models from cognitive science (ACT-R activation decay, Hebbian learning, forgetting curves) to decide what to remember and what to forget. Been running it in production for a month, 230K+ recalls. Just shipped v2 with multi-agent shared memory. https://github.com/tonitangpotato/engram-ai https://github.com/tonitangpotato/engram-ai-rust

Re: Ask HN: What Are You Working On? (March 2026)

#957
Recently, I've been working on Backdrop [1], a desktop app for taking and beautifying screenshots.

The idea came from trying to find a neat screenshot app on Linux for taking and sharing screenshots for another SaaS I'm working on: LanguagePuppy [2], but I couldn't find one that worked. So I rolled up my sleeves and started to make my own using Clojure.

The main features are: 1. Taking screenshots 2. Adding padding, shadow, and rounded corners to the screenshot 3. Adding background 4. Choosing the right ratio for your target social media 5. Adding watermark 6. And more

Eventually, it should work on Linux, Windows, and macOS when I launch.

P.S. I am building both with Clojure, a functional programming language, and Lisp.

[1] https://backdrop.whatacold.io/

[2] https://languagepuppy.com/

Re: Ask HN: What Are You Working On? (March 2026)

#958
I'm building Reveal, an open-source color separation engine for screen printing. It takes a full-color image and reduces it to 3-10 spot colors. This is not for CMYK 4-color process or simulated process ('halftones' using opaque colors).

The core is pure JavaScript with zero dependencies. It works in Lab color space (not RGB) using Wu quantization, CIE perceptual distance metrics, and a DNA fingerprinting system that matches each image to one of 20+ built-in archetypes (Golden Hour, Dark Portrait, Fine Art Scan, etc.).

There's a browser-based web UI that just needs Node.js — drop an image, get 4 different possible separations side by side, export as layered OpenRaster (for GIMP/Krita), or flat PNG, or layered PSD (for Photoshop). There's also a CLI and Photoshop plugin.

Apache 2.0: https://github.com/electrosaur-labs/reveal

Some of my screen prints: https://www.electrosaur.org

Post reply on HN