Anuj Rawat User

I am a prompt engineer with hands-on experience improving LLM outputs, developing prompt datasets, and fine-tuning models using RLHF. I have worked on annotation, caption quality assessment, and cross-functional model refinement.

At Softage AI I increased prompt response accuracy by 20% and designed prompt datasets for multiple models to enhance functionality. At NVIDIA I focused on image annotation and evaluating auto-generated captions to ensure accuracy and alignment with project goals.

I build data-driven tools and dashboards, including a Power BI churn dashboard that surfaced actionable insights such as a 27% overall churn and a 91% churn among customers with phone services. I leverage Python, Excel, and Power BI for analysis and visualization.

I bring strong attention to detail, communication skills, and a systematic approach to tracking issues and escalating them to senior stakeholders to continuously improve model performance and product outcomes.