Pitchlens watches hundreds of pitch videos so your judges can spend their time on the ones that matter. It scores each one the same way, shows exactly what was said to earn each score, and leaves the decision to people.
Upload videos in bulk, paste YouTube or Drive links, or open a public form where teams submit their own.
Every word is transcribed with its time — English, Hindi or both at once. Key frames show the slides and demos.
Eight dimensions, each scored several times against a written scale, each score backed by quotes checked against the transcript.
Judges watch, jump to the moments that matter, adjust any score with a reason, and see where they and the AI disagree.
A quote that is not in the transcript is thrown away. Every one that stays jumps the video to when it was said.
Each dimension is scored several times; how much they agree — not what the model claims — sets the confidence.
Accent, Hinglish and filler words never lower a score. Appearance is never assessed.
The AI is a first read. Judges score independently, blind if you choose, and their scores are what count.