Akshat Kumar

At Innodata India, I evaluated LLM outputs for quality, factual accuracy, reasoning, and instruction-following, contributing preference judgments and annotation reviews to RLHF workflows.

I've also performed QA across text, image, and audio AI tasks, identified generative AI failure patterns, and delivered actionable feedback to improve model performance. Through Excelerate × Saint Louis University, I developed project plans, budgets, contingency strategies, stakeholder communications, and QA processes for a global AI-focused virtual event.

I build practical AI tools, including a Streamlit app that converts business requirements into structured business-analysis documentation and an n8n job-aggregation workflow powered by Gemini AI and Google Sheets.