Recent submissions

Comment · 1 year ago
5 years ago, after struggling to land offers at FAANG companies, I started interviewing engineers who had successfully navigated the big tech interview process. My goal was to understand what it really takes to stand out and secure offers at the best companies.

I shared these insights online as a side project, and over time, it grew into a business—Onsites.fyi.

For the last 5 years, I’ve been collecting data from candidates about their interview experiences: companies they interviewed with, difficulty ratings, types of questions asked, time to offer, and more.

I’ve now built this into a searchable, sortable, filterable database so job seekers can get real insights into the onsite interview process at FAANG companies.

I wanted to get HN’s opinion on the tool, and if you find the data / information useful? Also, would love to hear what other data points you think we should be collecting, or features we should add!

The tool is built with Next, Postgres database and it’s hosted on Vercel.

Comment · 1 year ago
Hey HackerNews!

I’m super excited to introduce Parafact — a real-time, AI-powered factchecking platform designed to verify any writing (human or AI-written) using reliable sources from the web.

Problem

In a world of information overload, factchecking is more essential than ever. With information on the internet growing exponentially, it’s crucial to separate reliable sources from misinformation. Parafact helps bridge this gap, ensuring that the content you create or consume is both accurate and credible.

Who is this for?

For Content Creators : Write in our editor, click Factcheck, and instantly receive insights that help make your writing more trustworthy and identify any inaccuracies.

For Content Consumers : Found something online or someone has shared that you’d like verified? Just paste the content into our editor, and Parafact will instantly check it for accuracy.

We built Parafact to make real-time factchecking simple and accessible for everyone.

Our users come from a variety of fields — from writers and researchers to developers and lawyers — all of whom benefit from a tool that makes content verification easier and faster. For those looking to integrate factchecking directly into your own products (for content moderation or other needs), we also offer a developer API.

I’d love to hear your thoughts, answer any questions, and see how you might use Parafact! Learn more about Parafact here: https://parafactai.com/

We're also giving away everyone their first month free for the Essential Plan. Use the code `TRUTHSEEKING` during your checkout.

Comment · 2 years ago
Hi HN!

I recently discovered Perplexity.ai and I was amazed! I built a product this past weekend that I felt should just exist in the world.

With a knowledge engine such as Perplexity, a realtime fact checking platform is now feasible. DeepFact runs in the background as you speak and actively extracts claims and fact checks them.

The app is built using NextJS, Deepgram API(nova-2 model) for speech to text, and Perplexity API(llama-3.1-sonar-small-128k-chat) for real-time fact checking.

It is completely free, just add your own API keys to give it a try. There's also a simulated demo using Elon-Trump's interview on X if you just want to see how it works.

Ideally this product would be used to fact check any conversation in realtime: public speeches/presentations, lectures, meetings, debates, podcasts. The use cases seem endless.

There are times where it fails and hallucinates source URLs or it cites itself. There is plenty of room to improve the fact-checking algs.

Please try it out and I would love to hear any feedback or comments on how this could be used and ways to improve it. Thanks!

Comment · 2 years ago
Edit: Perplexity Pro users get $5 of free API credits per month.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton