Postingan

Menampilkan postingan dari April, 2025

Show HN: I made an AI Platform that gamifies applying to jobs https://ift.tt/w89Avei

Show HN: I made an AI Platform that gamifies applying to jobs Hi there! I've created ApplyNinja, a platform that uses AI to help jobseekers apply to jobs faster and better. What differentiates ApplyNinja from the other "auto-apply" platforms is the trained AI model behind it. It learns from your resume and the jobs you apply to, and it provides better and better results. Currently, it serves as a helper for the entire application process: - Track & Monitor job applications - Generates Resume suggestions and highlights the parts that are not so good - Auto-fixes and generates a new resume based on the suggestions - Creates Cover Letters based on your resume and the job description - Generates Company Insights, scraping user reviews from Glassdoor, Indeed etc. - Generates Salary Insights so jobseekers have a negotiation leverage - Generates Technical Interview Questions based on the Job Description On top of all of this, it provides a gamified experience, because consis...

Show HN: Experience iPhone's "Blob Keyboard" prototype from 2005 https://ift.tt/iJkZfEn

Show HN: Experience iPhone's "Blob Keyboard" prototype from 2005 Hi HN, I teach tech design history, and one of the key stories I cover is the development of the original iPhone keyboard by Ken Kocienda. Reading about it in his book "Creative Selection" is great, but I wanted my students (and now you!) to actually feel this step in the process. So, I built a web simulator of the "Blob Keyboard", Kocienda's very first attempt at a touchscreen keyboard that actually works, from September 2005: Try the Blob Keyboard: https://ift.tt/w62C3YA... - Tap for the middle letter - Swipe left or right for the side letters More on the github repo: https://ift.tt/k7EQxNl The Blob Keyboard prototype emerged during a UX crisis for iPhone team (their software keyboard just didn't work at all, fingers being too big, and the Newton failure loomed over them), highlighting how innovation is rarely a straight path. It was developed on a tethered touchscreen display c...

Show HN: Reharmonizing Racing in the Street by Bruce Springsteen https://ift.tt/sd8cVra

Show HN: Reharmonizing Racing in the Street by Bruce Springsteen I put together a short essay on re-harmonizing Racing in the Street by Bruce Springsteen. Just a simple demonstration of what’s possible today with a few hours and the right mix of technology https://ift.tt/STugxoE April 6, 2025 at 12:41AM

Show HN: I made an online free tool site https://ift.tt/QGcgNT3

Show HN: I made an online free tool site Hi HN! I've created [UFreeTools]( https://ift.tt/R2umoKN ), a collection of developer utilities (60+ tools and growing) that's completely free, ad-free, and runs entirely in the browser. No data leaves your device - everything is processed locally. ## Background As a developer, I found myself constantly searching for simple utilities like JSON formatters, UUID generators, and color pickers. Most existing options were cluttered with ads, required sign-ups, or sent data to servers. I wanted clean, fast tools that respected privacy. So I built UFreeTools as a modern SPA using Vue 3 and Vite with: - 100% client-side processing for all tools - Progressive loading with dynamic imports (only load what you use) - Dark/light theme support - Offline capability for most tools - Responsive design that works well on mobile ## Technical Details Some of the more technically interesting implementations: - *JWT Debugger*: Uses WebCrypto API for verificat...

Show HN: I made a platform to load and performance test API's https://ift.tt/BaboV7d

Show HN: I made a platform to load and performance test API's I built BuzzBench because I was frustrated with how complex performance testing tools have become. And I was ending up writing my own scripts to test endpoints and manually check out resource usage at the time of testing. Checkout demo: https://www.youtube.com/watch?v=yAnbZMoQvmQ https://buzzbench.io/ April 2, 2025 at 05:52PM

Muni Matters: Ridership on the Rise

Muni Matters: Ridership on the Rise By Erica Kato Year after year, we’re seeing a strengthened ridership recovery, reaching new highs in 2024, with a total of 158 million passenger trips, resulting in an increase of 13.5 million trips compared to 2023. The latest data highlights a significant increase in daily boardings and points to several high-performing routes. It also shows the impact of major events on Muni ridership -- people are using Muni more than ever to get to Warriors games, parades and street festivals, and so much more. Muni is the backbone of San Francisco’s economic vitality. Ridership grew on weekdays and weekends In 2024... Published April 04, 2025 at 07:00AM https://ift.tt/oOKtBTe

Show HN: Portowl – Simple Retro RPG https://ift.tt/4UiOnDj

Show HN: Portowl – Simple Retro RPG Experience first-hand the wonderful world of owl photography! Owls aren’t big on having their photo taken, so learning how may take some time. While you play, be on the lookout for powerful upgrades. https://ift.tt/AjvfKVc April 4, 2025 at 11:52PM

Show HN: Haystack Code Reviewer – Guided Tours for Pull Requests https://ift.tt/aeHPMnK

Show HN: Haystack Code Reviewer – Guided Tours for Pull Requests Hi HN! We’re excited to share Haystack Code Reviewer, a tool that analyzes, chunks, and organizes the changes made in a pull request and guides you through them in a logical sequence on an interactive canvas. We hope that Haystack really cuts down on the time that you spend piecing together what's happening in the pull request, so that you can spend time on the parts that matter, like providing feedback or internalizing the changes. As software engineers who’ve dreaded code reviews and always put them off as much as possible, we noticed how much time we spent poring over diffs to understand the broader picture or narrative of a pull request before even evaluating the code itself. Haystack Code Reviewer is designed to assemble that jigsaw puzzle for you, making code reviews less of a chore and more of a learning opportunity for both the reviewer and author! To play with demo code reviews, go to https://ift.tt/UQPIgWu a...

Uninterrupted Mobile While You’re Mobile: Cellular Service Comes to Muni Metro

Uninterrupted Mobile While You’re Mobile: Cellular Service Comes to Muni Metro By Melissa Culross Muni riders now can get cell service both above ground and inside our metro and subway tunnels. If you text or use your phone to browse the internet while riding Muni, you may have noticed that you now get cell service where you couldn’t before. Specifically, it’s likely that your signal no longer drops when the train you’re on is underground. Keeping you connected while you ride Cellular service has been available in the Central Subway since it opened in 2023. And now, we have service in Muni Metro from the Embarcadero to West Portal Station. Service also is being added to the Sunset Tunnel... Published April 03, 2025 at 07:00AM https://ift.tt/X0eQhnq

Show HN: LocalScore – Local LLM Benchmark https://ift.tt/3rHkZTw

Show HN: LocalScore – Local LLM Benchmark Hey Folks! I've been building an open source benchmark for measuring local LLM performance on your own hardware. The benchmarking tool is a CLI written on top of Llamafile to allow for portability across different hardware setups and operating systems. The website is a database of results from the benchmark, allowing you to explore the performance of different models and hardware configurations. Please give it a try! Any feedback and contribution is much appreciated. I'd love for this to serve as a helpful resource for the local AI community. For more check out: - Website: https://localscore.ai - Demo video: https://youtu.be/De6pA1bQsHU - Blog post: https://ift.tt/r6nqylA - CLI Github: https://ift.tt/ERAUPMX... - Website Github: https://ift.tt/REscDMP https://ift.tt/UnE8wTs April 3, 2025 at 11:32PM

Show HN: Synto - micro-learning app to help you deeply understand classic books https://ift.tt/T0un9ho

Show HN: Synto - micro-learning app to help you deeply understand classic books https://ift.tt/iPjRu8J April 3, 2025 at 11:14PM

Show HN: Arrakis – Open-source, self-hostable sandboxing service for AI Agents https://ift.tt/3oAZFME

Show HN: Arrakis – Open-source, self-hostable sandboxing service for AI Agents Hi HN! My name is Abhishek. I've spent my career working on Operating Systems and Infrastructure at places like Replit, Google, and Microsoft. Today, I'm excited to launch Arrakis: an open-source and self-hostable sandboxing service designed to let AI Agents execute code and operate a GUI securely. GitHub: https://ift.tt/LwFc5NJ Watch Claude build a live Google Docs clone using Arrakis via MCP – with no re-prompting or interruption. https://www.youtube.com/watch?v=IZ5cAnhAdPQ Key Features - Self-hostable: Run it on your own infra or Linux server. - Secure by Design: Uses MicroVMs for strong isolation between sandbox instances. - Snapshotting & Backtracking: First-class support allows AI agents to snapshot a running sandbox (including GUI state!) and revert if something goes wrong. - Ready to Integrate: Comes with a Python SDK ( https://ift.tt/KNMqvge ) and an MCP server ( https://ift.tt/2IHytmF )...

Show HN: Agent File (.af) – An open file format for agents https://ift.tt/tOTDAgV

Show HN: Agent File (.af) – An open file format for agents Hi HN - We’re building Agent File (.af), which makes it possible to re-create the exact same agent (including memories, tools, message history, configs, etc.) across different machines. A big difference between LLMs and agents is that agents have associated state: system prompts, editable memory (personality and user information), tool configurations (code and schemas), and LLM/embedding model settings. While you can run the same LLM as someone else by downloading the weights, there’s no “representation” of agents that allows you to re-create an instance of an agent across services. Agent File (.af) is an open standard file format for serializing stateful agents. Originally designed for the Letta framework, .af is a human-readable representation of all the associated state of an agent to reproduce the exact behavior and memories. To demonstrate .af, we also made a few example agents with download links to .af: - MemGPT: An agen...

Show HN: Mermaid Chart VS Code Plugin: Mermaid.js Diagrams in Visual Studio Code https://ift.tt/5SMqlyi

Show HN: Mermaid Chart VS Code Plugin: Mermaid.js Diagrams in Visual Studio Code https://ift.tt/rFQeX3W April 2, 2025 at 11:33PM

Show HN: A Chrome extension to give you back control over short-form videos https://ift.tt/ExRdDHB

Show HN: A Chrome extension to give you back control over short-form videos Hi HN! I built this little extension to prevent, in my opinion, the most offensive anti-pattern used by tech companies. That is removing the seek bar in short-form videos. The "seek bar" is the bar at the bottom of a video that progresses as you play the video, and that you can click on or drag to skip around. Why companies ever thought it was a good idea to get rid of this I don't know, but I find it infuriating, so I decided to add it back for myself and thought others might like it too. ReelControl adds a progress bar and seeking capabilities to videos on Instagram, YouTube Shorts, and Facebook Reels. I do sometimes enjoy watching short-form content and I've found that with this extension enabled I can be more mindful about it and get sucked in way less. I'm also on my phone less because I tend to favor the web versions of these platforms now. Open source--PRs and issues welcome! https:...

Show HN: Qwen-2.5-32B is now the best open source OCR model https://ift.tt/h7KTyvu

Show HN: Qwen-2.5-32B is now the best open source OCR model Last week was big for open source LLMs. We got: - Qwen 2.5 VL (72b and 32b) - Gemma-3 (27b) - DeepSeek-v3-0324 And a couple weeks ago we got the new mistral-ocr model. We updated our OCR benchmark to include the new models. We evaluated 1,000 documents for JSON extraction accuracy. Major takeaways: - Qwen 2.5 VL (72b and 32b) are by far the most impressive. Both landed right around 75% accuracy (equivalent to GPT-4o’s performance). Qwen 72b was only 0.4% above 32b. Within the margin of error. - Both Qwen models passed mistral-ocr (72.2%), which is specifically trained for OCR. - Gemma-3 (27B) only scored 42.9%. Particularly surprising given that it's architecture is based on Gemini 2.0 which still tops the accuracy chart. The data set and benchmark runner is fully open source. You can check out the code and reproduction steps here: - https://ift.tt/gKB8lQW... - https://ift.tt/sILXZny - https://ift.tt/tDIR1xL https://ift.tt...

Show HN: Generate Anki Decks from YouTube Videos to Anki Converter https://ift.tt/OuUcE1N

Show HN: Generate Anki Decks from YouTube Videos to Anki Converter Hello! I love anki and have been using it for years to learn mandarin, but I'm lazy and always struggled to get over the "motivation hump" to sit down and create a deck. Recently I've been consuming a lot of AI content from youtube, and spent 3 days creating a deck for Andrej Karpathy's "Deep Dive into LLMs" video. This is impractical - so I built a little youtube to anki converter @ asimpleai.com It's rough around the edges... but I would like it to eventually become something useful to community. Feedback and suggestions are welcome! peace https://asimpleai.com/ April 1, 2025 at 11:17PM