About ReelsCutter

We Help Creators Work Smarter

ReelsCutter was born from a simple frustration: great long-form content deserves to be discovered. We built AI that finds those moments so creators spend less time scrubbing timelines and more time creating.

Our Mission

Every long-form video has 3–7 moments that could go viral. Our mission is to surface those moments instantly using AI — so every creator, regardless of editing skill or resources, can compete at the highest level of short-form content.

How the AI Pipeline Works

Our Python backend processes your video through a 4-stage pipeline, all running on GPU-accelerated infrastructure.

Step 1

Audio Extraction

yt-dlp fetches the video and ffmpeg extracts the audio track at optimal quality.

Step 2

Transcription

OpenAI Whisper (large-v3) or AssemblyAI transcribes with word-level timestamps.

Step 3

AI Analysis

Our LLM pipeline scores each segment for virality signals: hooks, quotability, story arc.

Step 4

Structured Output

Returns JSON: {start_time, end_time, transcript, virality_score, reason} per clip.

// API Response Shape

{
  "clips": [
    {
      "start_time": 124,      // seconds
      "end_time": 189,
      "transcript": "The biggest mistake most founders make...",
      "virality_score": 94,  // 0-100
      "reason": "High emotional hook with personal failure story"
    }
  ]
}

Meet the Team

RG

Rajan Gupta

Founder & CEO

Former ML engineer. Built content tools used by 500K+ creators.

MK

Meera Krishnan

Head of AI

PhD in NLP. Previously at DeepMind. Obsessed with language models.

AB

Aryan Bose

Lead Engineer

Full-stack + Python. Built the processing pipeline from the ground up.

TM

Tanya Mishra

Growth

Content creator turned growth hacker. 200K subs before joining ReelsCutter.

Want to Join Us?

We're a small, remote-first team solving real problems for creators. Always hiring great people.