Data Annotator/Quality Analyst at OTAKU.HUGO COMPANY LIMITED (2023-06 – Present)
Data annotation and quality analysis role focused on achieving high accuracy in model training and ensuring quality requirements adhere to company policies.
- Achieved 99.5% accuracy by completing data annotation tasks accurately and effectively.
- Collection of all the essential statistics required and identify the root cause, analyzing methods to identify the defect.
- Achieved trained model with high precision through effectively annotating data with 99% accuracy.
- Creation of plans to ensure that quality requirements adhere to the company's policies.
- Assisted clients with the process of tagging and setting guidelines
- Auditing the job process intermittently to ensure error-free procedures.
- Providing knowledge transfer to new teammates.
- Providing efficient solutions for various types of issues.
- Evaluation of photorealism and visually appealing pictures with error tagging.
- Conducting benchmark quality comparisons and standalone evaluation.
- Detection of subtle errors, unnatural faces, and AI generated images.
Image Annotation & AI Evaluation Specialist at ABAKA AI (2025-10 – Present)
Image annotation and AI evaluation specialist working on AI vision and multimodal model training with focus on quality dimensions and anatomical accuracy.
- Performing image editing, sampling, and resampling tasks to support AI vision and multimodal model training.
- Evaluating images across key quality dimensions including Instruction Following, Image Consistency, Visual Quality, and AI-ness.
- Working on Human Anatomy & Physical Plausibility projects, identifying anatomical inaccuracies, unrealistic body proportions, incorrect poses, and motion inconsistencies.
- Rated and reviewed photorealistic images, synthetic imagery, and infographics for realism, coherence, and compliance with prompts.
- Conducted RSHF (Resampling with Human Feedback) by selecting, correcting, and refining image outputs to improve downstream model learning.
- Flagging subtle visual defects such as unnatural faces, distorted limbs, lighting inconsistencies, texture artifacts, and background anomalies.
- Supported image generation benchmarking, comparative evaluation, and quality assurance across multiple AI model outputs.
- Ensuring strict adherence to annotation guidelines and delivered consistent, high-quality evaluations across diverse image datasets.
Data Annotation Expert (LLM / MMLLM) at Mindrift (2025-05 – Present)
Data annotation expert specializing in Large Language Model and Multimodal LLM training and evaluation with focus on response quality assessment and RLHF workflows.
- Worked extensively on Large Language Model (LLM) and Multimodal LLM (MMLLM) training and evaluation projects.
- Annotated, reviewed, and evaluated model responses for tasks including instruction following, question answering, summarization, conversational AI, content moderation, and safety alignment.
- Applied strict annotation guidelines to assess accuracy, helpfulness, honesty, tone, formatting, insightfulness, and clarity of AI-generated responses.
- Identified and flagged hallucinations, reasoning errors, bias, outdated information, and policy violations in model outputs.
- Performing response ranking and preference selection as part of Reinforcement Learning from Human Feedback (RLHF) workflows.
- Wrote high-quality reference answers to guide model optimization and improve real-world performance.
- Maintaining high consistency and precision while working on large-scale datasets under tight quality thresholds.
- Collaborating with quality teams to improve model reliability, alignment, and user satisfaction.
Cloud & Automation Support Specialist at Lionbridge (2025-08 – Present)
Cloud and automation support specialist providing infrastructure support and automation tasks using Python, Microsoft Azure, and Amazon Web Services.
- Supported cloud-based infrastructure and automation tasks using Python, Microsoft Azure, and Amazon Web Services (AWS).
- Assisting in setting up cloud-based development environments
- Monitoring resource usage and optimized cloud performance
- Implementing access controls and basic security configurations
- Assisted in deploying web-based applications to AWS infrastructure
- Monitoring resource utilization and applied cost-optimization practices
- Supporting basic cloud networking and security configurations
- Integrated Python scripts with cloud resources for automation and monitoring
- Assisted in debugging and optimizing existing Python scripts
