Adrija Vishwakarma

AI Data Annotator and Model Evaluation Specialist with hands-on experience across platforms including Appen, Telus International, Outlier.ai, Labelbox, and Alignerr. I work on image annotation (bounding boxes, polygons, UI-button labeling), text and speech evaluation, and LLM output quality rating, with a strong focus on accuracy and consistency.

I have completed AI response evaluations and annotations, covering UI datasets, multilingual prompts, tone/emotion assessments, function-calling verification, and safety/content checks. I’m experienced in identifying hallucinations, ranking model outputs, and evaluating prompt-based responses.

Alongside professional work, I’ve built projects in Machine Learning, Computer Vision, NLP, and IoT, including fraud detection systems, CNN image classifiers, spam detection models, and generative AI text generators. My technical foundation includes Python, Scikit-learn, TensorFlow/Keras, basic data preprocessing, and GitHub workflow.

I hold a B.Tech in Electronics & Communication Engineering (Honours in IoT) with a 9.03/10 CGPA, and certifications from IIT Delhi, SkillVertex, Maven Silicon, and Kaggle. I’m passionate about high-quality dataset creation and improving LLM performance through structured evaluation and precise annotation.