Sourabh Dabadiya

Extracted problem-solution pairs from STEM educational resources to enhance dataset quality.

Transcribed mathematical equations and solutions in LaTeX for increased accuracy.

Contributed to development of AI models, improving accessibility of educational resources.

• Labeled and annotated text/image data for AI projects.
• Collaborated with QA teams to maintain high documentation standards.

• Developed and assessed mathematical reasoning prompts in algebra, calculus, probability, and applied mathematics.
• Created LaTeX-based solutions that ensured clarity and pedagogical effectiveness.
• Conducted adversarial testing of LLMs by generating complex applied math problems.
• Performed quality assurance reviews of AI responses for accuracy and logical consistency.
• Contributed to benchmarking datasets for assessing mathematical reasoning and real-world applications.

• Annotated datasets in STEM and language domains for AI training.
• Ensured correctness in grammar, logic, and technical solutions.

• Reviewed and solved advanced math and computer science problems to validate AI outputs.
• Designed domain-specific STEM questions to enhance AI reasoning performance.
• Delivered structured, step-by-step solutions with academic rigor and clarity.
• Provided critical feedback on AI-generated responses for correctness, depth, and clarity.