Farhan Q

Analyst AI/LLM Practices

Innodata Inc • Temps plein

Oct 2024 - Jan 2026 • 1 yr 3 mos

● Conduct comprehensive text, image, and video annotation for diverse clients globally, ensuring delivery of high-quality, training-ready datasets for generative AI and large language models. ● Maintain impressive accuracy rate of 90%+ across all annotation projects, significantly enhancing client satisfaction and model performance metrics. ● Apply advanced annotation techniques tailored to specific client requirements and project objectives, optimizing dataset quality and outcomes. ● Execute meticulous data cleaning and preprocessing to ensure datasets are production-ready for generative AI, machine learning model training, and deployment.

ML Data Associate

Amazon • Temps plein

Oct 2018 - Mar 2023 • 4 yrs 5 mos

● Collaborated on 8 diverse image and video annotation tracks for Amazon GO (world's most advanced computer vision-based checkout technology), consistently achieving 90%+ accuracy across all deliverables. ● Provided ground truth validation and annotation for critical datasets, ensuring data accuracy and integrity for downstream machine learning model training and deployment. ● Conducted thorough root cause analysis on ML model errors, identifying data quality issues and implementing corrective measures to improve model performance and reliability. ● Executed comprehensive data preprocessing and cleaning workflows to enhance machine learning model performance and reduce prediction errors. ● Collaborated closely with cross-functional stakeholders and ML engineers to deliver high-quality golden datasets critical for production model deployment.