• *Role Overview:** Mercor is partnering with a leading AI lab to improve current AI models by creating timestamped descriptions of video clips. Experts will do the following: 1. Watch short videos, no longer than 2-5 minutes 2. Break down the video into key segments, mark down the start + end time of each segment, as well as provide a brief summary of what is happening in each segment 3. For each segment, identify the most important video frame / frames

• 4. Identify the most important frame in the entire video, and provide a brief summary of what is happening in that frame
• Ideal Qualifications:**

- Are very detail-oriented - Excel at following precise instructions and quickly adopting new requirements - Can apply judgment to decide which segments in a video are most important, and what the key frame(s) are in a video - Very comfortable with watching + understanding videos from major social media platforms (such as Facebook, TikTok, Instagram Reels, and YouTube Shorts) - Possess strong written and verbal communication

• Fluent in English
• More details:**

- You’ll collaborate directly with Mercor’s project leads. - Start dates are rolling; we aim to onboard qualified experts within a few days (1‑2) of application. - Up to 40 hours of work available per week.

• The project is expected to last a minimum of 4-5  weeks with potential for extension.
• Screening Process:**

- You will need to complete a short interview and form, which will take a total of 30-50 minutes. - Currently only accepting applicants from the U.S., UK, and Canada.