At INNODATA, I analyze Large Language Model outputs to identify accuracy gaps, bias patterns, and failure cases across diverse prompts.
I perform structured data validation and prompt-level analysis to support response quality, consistency, and policy compliance.
I apply defined quality and safety frameworks to support large-scale LLM evaluation and reduce incorrect or non-compliant outputs.
I also created a Python-based student database management system with SQL, and simulated a Si-Ge pocket vertical TFET for low-power device applications.
