Re:Ask HN: What Are You Working On? (June 2025)
by noisy_boy on Sunday June 29, 2025 @09:55PM
(#44418438)
I think if you allow a set of YouTube videos as input, it'll be quite powerful coupled with transcription ability of LLMs. Lots of people consume content that way. As an added bonus, you can show the performance summary about the sections the user did well or not so well on with video links to those timestamps for them to go back and review.
Re:Ask HN: What Are You Working On? (June 2025)
by whitefang on Monday June 30, 2025 @12:29AM
(#44419385)
I'm building an AI for Customer Support.
Here's the summary: - read all your sources - public websites, docs, video - answer questions with confidence score and no hallucinations with citations - cut support time and even integrates directly into your customer facing chatbots like Intercom
Still deliberating on the business model. If anyone would be interested in taking a look, I would love to show you.
Started as a very simple app for me to play around with OpenAI’s API last year then morphed into a portfolio project during my job search earlier this year. Now happily employed but still hacking on it.
Right now, a user can create a quiz, take a quiz, save it and share the quiz with other people using a URL.
Demo: You can try out the full working application at https://quizknit.com
Github Links: Frontend: https://github.com/jibolash/quizknit-react , Backend: https://github.com/jibolash/quizknit-api