Video & Image Annotation Specialist - TJ / 1872675

Our client Opsbotree is looking for a Video & Image Annotation Specialist in Karachi

Location: Razi Road, PECHS, Block 6, Karachi (on site applications only)

We are building a dedicated Core Annotation Team to support the tagging of marketing assets (images and videos) inside our own annotation platform. This team will replace a previous outsourced annotation workflow, moving the work fully in-house through our proprietary tool.

As a Core Team Annotator, you will be one of the founding members of this function. You will not just tag assets — you will become an expert in the annotation methodology, help train and onboard additional annotators as the team scales, and act as a key point of contact between the annotation workforce and our data team on technical and edge-case issues.

Responsibilities

• Annotate marketing assets (static images and videos) by tagging defined Areas of Interest (AOIs), including: headline, brand, copy, call-to-action (CTA), and product placement, using our in-house annotation platform.
• Follow detailed annotation instructions and guidelines provided by the data team, applying them consistently and accurately across large volumes of assets.
• Conduct peer quality assurance (QA), reviewing other core team members' work against the annotation instructions and flagging inconsistencies (as the team scales).
• Identify and escalate edge cases or ambiguous assets that fall outside the current guidelines, documenting them clearly for review and resolution by the data team.
• Help onboard, train, and support members of the scalable (surge) annotation team when volume increases, ensuring they understand tagging standards and platform usage.
• Maintain annotation speed and throughput targets while upholding accuracy standards.
• Provide feedback on annotation instructions, edge cases, and platform usability to help continuously improve the process.
• Act as a liaison between the annotation workforce and the data team on day-to-day technical or process questions.