Ramsha Hasib

AI LLM Analyst

Innodata

Apr 2026 - Jul 2026

Engaged on the Superintelligence Lab line of business (LOB), supporting large-scale LLM data operations. Maintained annotation quality benchmarks with 90%+ inter-annotator agreement across AI/LLM training datasets, ensuring data consistency and reducing downstream bias and post-deployment error correction. Designed and developed scalable data models and interactive Power BI dashboards, reducing reporting time by 30% and improving visibility into business KPIs. Gathered reporting requirements from stakeholders and cross-functional teams, translating business needs into KPI dashboards and structured reports. Performed data validation, annotation review, and preprocessing on large-scale, high-quality datasets for AI and Large Language Model (LLM) applications, ensuring dataset accuracy and model readiness. Identified data quality issues across datasets of 1–2 million records and implemented corrective actions, reviewing output from a team of 100 annotators and improving data accuracy to 95%+, enhancing AI model training outcomes. Transformed large, structured datasets using Python, SQL, and ETL processes, ensuring data accuracy and readiness for AI/LLM training and evaluation workflows.