It would take a lot to setup an API for certain resources when you consider stuff from security to compliance, but if you could just slap on a React plugin on already existing UIs so now agents can use the same site as humans already do.
It is becoming increasingly clear that AI agents will do more on the internet and operate many websites, so why are web UIs still built only for how humans use them? Why isn't it being extended so AI agents can talk to websites? Right now, the ridiculous amount of complexity it takes for AI agents to operate a website and perform simple operations is crazy; look at the Claude Chrome extension,...
If your team is hiring and conducting interviews, I urge you to consider ScreenStack: an all-in-one interview platform that replaces your fragmented stack of tools.It was built by engineers who have spent years at FAANG, conducted over 400 hours of interviews, and felt the pain firsthand: juggling Zoom, CoderPad, Excalidraw, and an AI notetaker, then burning hours stitching them together to...
I've spent almost a decade at one of the FAANG companies, and I've conducted over 400 hours of software engineering interviews. The frustration led me to build ScreenStack.You see, the interviewing stack today is fragmented. To conduct an interview, you'll have to schedule a conference call on Zoom/Google Meet/MS Teams, then have a CoderPad session, and, most recently, connect your AI...
I built Moments to get this game idea out of my head, finally. My original goal was to run on-device models specifically in mobile browsers, but running local vision models directly in phone browsers is still very much too early, so I focused on desktop.How it works:- You upload a photo.- A local vision model running entirely in your browser captions it and picks a prominent object from the...
I’m trying to understand the rationale behind Anthropic limiting certain types of third-party usage, e.g., OpenClawFrom a naive perspective, more usage should mean more revenue since customers pay per token. So why restrict it? If a third party is generating heavy usage, isn’t that ultimately beneficial for revenue and growth?What other factors are they considering that are not immediately...
For interviews, I search company-specific tags to find discussions and tips, ChatGPT makes it even easier now to extract insights from those threads. I also use it for compensation research. Occasionally, I’ll check company gossip, and I have to say, Blind has correctly predicted layoffs at my employer twice. Beneath all the noise, some people really do share valuable inside information.
That’s why I treat Blind as a data-gathering tool, not a hangout. I mainly open it for interviews, negotiations, compensation benchmarks, or to get the general sentiment around a company. Honestly, I wish they had an API like Reddit’s, it would make pulling insights so much easier.
Teamblind has been so impactful for me that I’m more than happy to pay for it. While there’s certainly noise on the platform, I’ve learned to focus on the insightful conversations and resources that matter. If you are in the US it might not be so useful, but for us outside, it is gold.
I've been on HN for way longer than Blind, but Blind had had the most impact in my career & I'm grateful for that.
I work at a FAANG and have spent the last 5+ years building video conferencing tools. I hacked together a Chrome extension to fix my video colour. I planned to add more features: local recordings, on-device AI transcripts and summaries, and clean screenshots of screen shares. Then my company shut down our in-house tool and moved to Zoom. Around the same time, I joined a new team and had to...
https://github.com/dokasto/Saidia
Saidia is an offline-first AI assistant tailored for educators, enabling them to generate questions directly from source materials.Built using Electron, Ollama, and Gemma 3n, Saidia functions entirely offline and is optimised for basic hardware. It's ideal for areas with unreliable internet and power, empowering educators with powerful teaching resources where cloud-based tools are impractical...
Hi HN,I’ve been working on a game idea inspired by Wordle, but with a unique twist: it uses your own photos to generate guessing words. Here’s how it works: the app picks a random picture from your gallery. It uses a small language model (SLM), running entirely on your phone, to identify a word from the image. The chosen word could describe an object, the mood, or any notable feature in the...
How so? I have a lot of respect for ML engineers and I know what they do is no easy feat. However, I do not intend to get into ML, I see it as a tool to help me achieve an objective. My goal is to utilise said tool to achieve an objective, I have no intention of becoming an expert in this area. It's just like CSS when you want to use it, do you have to care about how the browser rendering works? No you don't, you simply use CSS.
I am an experienced software engineer with decent knowledge of Python. I've used Machine Learning Models in the past to do stuff like background blurring etc.What do I want? I want to be able to train models and use them for the applications I intend to build. I honestly do not care about the Math, statistics or theory behind them. I just want to know enough to be able to train a model, use a...