We're planning on keeping it free - enjoy and let us know if there's anything else that comes up!

Comments URL: https://news.ycombinator.com/item?id=42139531

Points: 4

# Comments: 0

The Problem: There’s a ton of valuable content on the internet, but finding the “best parts” of long videos is frustratingly inefficient. Current video search methods haven’t evolved much since the early days of YouTube and aren’t designed for today’s massive volume and variety of information. Instead, we’re left scrubbing through long videos or, worse, missing valuable content entirely due to decision paralysis. Remy’s goal is to offer a smarter way to surface exactly what you’re looking for, in a fraction of the time.

The Solution: In less than a minute, Remy delivers custom playlists that isolate the best video moments from across the internet. It finds, clips, and organises segments to get you exactly what you need — without the endless search and skip game.

How it works:

Remy is powered by a stack of LLMs (and some non-LLM magic) designed for fast, focused video search and transcript processing. Here’s the pipeline:

1. Request Analysis When you send a message, the system decides whether to provide an immediate response, search the web, or start assembling video clips. If video is the best option, Remy generates a playlist outline with concise titles and detailed descriptions tailored to your query. For temporal queries, Remy automatically adjusts context to absolute dates (e.g., “tomorrow’s NBA games” → “November 15th NBA Games”).

2. Content Retrieval Using a ‘wide net’ approach, Remy generates a large set of targeted queries and searches the web for videos that could match your needs.

3. Multi-Step Filtration & Processing Each video goes through:

- Transcript Pull: Extracts YouTube transcripts.

- Non-LLM Filter: Filters out low-quality or AI-generated content based on YouTube stats, creator channels, release date, and other parameters.

- Punctuation Restoration (BERT): Restores punctuation for better LLM comprehension.

- Clipping: Uses an LLM to locate and clip the most relevant segments of the transcript based on your request.

- Evaluation: Uses an LLM to score clips on relevance, completeness, and interest level. Only the best make it to the final playlist.

4. Reordering & Overview A final LLM gets fed the top 16 clips for each topic, filters them down to the best 4 (at most), and arranges them for maximum uniqueness. Each section gets a brief overview, with added context pulled and cited from the web to give you additional context.

The result? A playlist tailored to your exact query, delivered in under a minute, without unnecessary noise.

We’d love to hear your thoughts and feedback. If there’s something you’re curious to try or any edge cases that it's not handling, let us know. And if you come across any bad clips, please use the report button to flag them! Thanks

https://www.useremy.com/

Comments URL: https://news.ycombinator.com/item?id=42139277

Points: 2

# Comments: 2

The Problem: There’s a ton of valuable content on the internet, but finding the “best parts” of long videos is frustratingly inefficient. Current video search methods haven’t evolved much since the early days of YouTube and aren’t designed for today’s massive volume and variety of information. Instead, we’re left scrubbing through long videos or, worse, missing valuable content entirely due to decision paralysis. Remy’s goal is to offer a smarter way to surface exactly what you’re looking for, in a fraction of the time.

The Solution: In less than a minute, Remy delivers custom playlists that isolate the best video moments from across the internet. It finds, clips, and organises segments to get you exactly what you need — without the endless search and skip game.

How it works:

Remy is powered by a stack of LLMs (and some non-LLM magic) designed for fast, focused video search and transcript processing. Here’s the pipeline:

1. Request Analysis When you send a message, the system decides whether to provide an immediate response, search the web, or start assembling video clips. If video is the best option, Remy generates a playlist outline with concise titles and detailed descriptions tailored to your query. For temporal queries, Remy automatically adjusts context to absolute dates (e.g., “tomorrow’s NBA games” → “November 15th NBA Games”).

2. Content Retrieval Using a ‘wide net’ approach, Remy generates a large set of targeted queries and searches the web for videos that could match your needs.

3. Multi-Step Filtration & Processing Each video goes through:

- Transcript Pull: Extracts YouTube transcripts.

- Non-LLM Filter: Filters out low-quality or AI-generated content based on YouTube stats, creator channels, release date, and other parameters.

- Punctuation Restoration (BERT): Restores punctuation for better LLM comprehension.

- Clipping: Uses an LLM to locate and clip the most relevant segments of the transcript based on your request.

- Evaluation: Uses an LLM to score clips on relevance, completeness, and interest level. Only the best make it to the final playlist.

4. Reordering & Overview A final LLM gets fed the top 16 clips for each topic, filters them down to the best 4 (at most), and arranges them for maximum uniqueness. Each section gets a brief overview, with added context pulled and cited from the web to give you additional context.

The result? A playlist tailored to your exact query, delivered in under a minute, without unnecessary noise.

We’d love to hear your thoughts and feedback. If there’s something you’re curious to try or any edge cases that it's not handling, let us know. And if you come across any bad clips, please use the report button to flag them! Thanks

Comments URL: https://news.ycombinator.com/item?id=42136968

Points: 6

# Comments: 6

Anecdotally, myself and every engineer I know has transitioned over the last 6 months from never using Claude and it being discussed as unusable (relative to GPT 4o) to everyone now using Claude while ChatGPT (specifically o1-preview) increasingly feels like it has become unusable for coding. 3.5 Sonnet seems to have been the big turning point.

Yet, all I see online is about how o1 is such an insane step forwards. I just don't see or feel it - in fact actually I vehemently disagree.

Would love to hear what you use and if that's changed at all - and if so, why so.

Sincerely, baffled

Comments URL: https://news.ycombinator.com/item?id=42129187

Points: 2

# Comments: 4

For whatever you search, Remy finds the best videos from across the internet then compiles the essential parts into custom playlists, saving you hours.

Too much time is wasted scrubbing through long videos looking for the parts you’re interested in and too many videos are missed entirely due to decision paralysis. In less than a minute, Remy will cut through the noise for you.

Ask: please try it and be critical, thanks.

Comments URL: https://news.ycombinator.com/item?id=42129135

Points: 2

# Comments: 0

The Problem:

Remy was born out of a simple frustration I had: after taking the Gordon Ramsay Masterclass, I realised I had already seen most of the content scattered across different YouTube videos over the years.

However, after going back to scouring YouTube to compile all of the videos I was interested in, I found that most of what I was looking for was buried in much longer videos. So, the question became: Why isn’t there one place to find every clip I’m looking for, no matter what video it’s hidden in? From that, Remy began to take shape.

The Solution:

We’ve since developed fast, powerful video search and understanding models, designed to save you time and effort by finding you exactly what you asked for, no matter how niche or specific. Remy searches across billions of videos in seconds, letting you iterate and hone in on the clips that matter most to you.

With real-time clipping, Remy looks for the best clips for your search every time, not just the closest clip we have.

How Remy Works:

Clipping long-form videos into short-form content isn’t new—tools like Opus, Klap, and Veed.io have helped creators do this for years. However, the clips that creators select to maximise engagement often don’t align with what you’re specifically looking for - but the content you want is still in those videos! We’re introducing consumer-focused clipping, tailored to what you want to see, for the first time.

Remy combines a chat interface with our video search engine. Just tell Remy what you’re looking for—whether it’s a specific person discussing something, a topic, or a comparison—and it will pull up the precise moments you asked for, from YouTube videos, interviews, podcasts, etc.

The result? A seamless experience where you’re only watching what interests you, and finding it in a fraction of the time.

Why Remy is Different:

I know, chatbots are feeling a little overdone - but sometimes they can be useful and we think that’s the case here. Remy’s chatbot form combines conversational and iterative search, powering focused rabbit-holes.

Remy doesn’t just search over an index of videos that we maintain, it also has web search capabilities to get the most up to date videos as well. We use Remy to ask questions about changes in San Francisco politics, recent sports games, and new music that’s coming out.

As your searches evolve, Remy understands the context and narrows down results in real time, finding clips that otherwise might be buried deep within hours-long videos. Have questions about a clip it found? Just ask in the chat and get answers rooted in the content Remy found for you.

Some Of Our Favourite Queries:

“The history of Le Gavroche and the Roux brothers.” “Stories about the writing of The Lord of the Rings” “Tell me about a sea animal I probably didn’t know existed”

What You Can Do with Remy:

Catch up on current events - “What’s going on around the world today?” Explore personal stories - “Gordon Ramsay talking about his relationship with Marco Pierre White.” or “Mark Zuckerberg talking about the impact of social media influencers” Make product comparisons - “Show me the differences between Oncloud Ultra 2s and Hoka Speedgoat 6s … Show me people just talking about the pros of the Hokas over the Onclouds … Now just people talking about the tread pattern of the Hokas.” Trace artistic influences - “How did A.G. Cook influence hyperpop?”

The Bottom Line:

Remy finds the exact moments you’re searching for, no matter where they are in a video, or how deep Google’s SEO has hidden them. Spend less time sifting through content and more time getting to the clips that really matter. With Remy, the perfect clip is just a chat away.

Try it out and let us know what you think! Thanks

Comments URL: https://news.ycombinator.com/item?id=42054777

Points: 2

# Comments: 0