Role overview:
You'll focus on ensuring the quality of AI data, Computer Vision/Vision and Language Models (VLM) prior to their deployment in real-world products.
The role is significant for building a robust, accurate and scalable data-model pipeline, supports AI and Robotics systems.
Responsibilities:
Ensure the quality of multimodal datasets (image, video, text, captions, instructions) for VLM training and evaluation, collaborating with internal teams and external data vendors.
Define and enforce data quality standards for VLMs, covering visual accuracy, textual correctness, vision-language alignment, semantic consistency, and bias/noise detection.
Review and audit data pre-and post-annotation, identifying systemic issues (like misaligned captions, hallucinated descriptions, weak prompts and inconsistent labeling. )
Design evaluation datasets, test protocols for VLM use cases including VQA, instruction following, multimodal reasoning, and human-robot interaction scenarios.
Evaluate VLM performance through quantitative metrics and qualitative behavior analysis, focusing on correctness, consistency, robustness, latency, and stability in real deployments.
Conduct error analysis and provide clear recommendations for data improvement, prompt refinement, model retraining, and release (go/no-go) decisions.
Qualifications:
Minimum 3+ years of experience, strong analytical mindset with high attention to data quality and model behavior.
Experience working with large-scale datasets and structured evaluation workflows.
Experience working with AI data pipelines, Computer Vision, VLMs, or multimodal datasets.
Hands-on experience with annotation or review tools (e.g.
Label
Studio, CVAT, or custom review systems).
Experience working with external data vendors or crowdsourced annotation teams.
Familiarity with evaluating AI systems beyond raw metrics, focusing on semantic correctness and user-facing behavior.
Solid understanding of how modern VLMs work (vision encoder + language model, alignment, prompting).
Collaborate with
AI engineers, researchers, and data vendors.
Ability to read, understand technical documentation and research papers in English.
Preferred Qualifications:
Basic Python skills for data analysis/evaluation scripting.
Understanding of MLOps / model lifecycle management.
Exposure to robotics, embodied AI, or human-robot interaction use cases.
Personality/ Attitude
Proactive, dedicated, business-oriented, responsible and willing to learn.
Good communication skills, creative problem-solving skills and attention to detail.