At Ethara AI, I evaluated LLM responses for quality, relevance, and instruction-following. I also created and refined prompts, instructions, and evaluation datasets.
I contributed to fine-tuning and post-training workflows through annotated data preparation, model evaluation, and benchmarking, and analyzed responses to identify quality issues and inconsistencies.
I developed Prepin, an AI-powered interview preparation web app with personalized questions and career guidance using Gemini AI. My projects also include a machine-learning crowd counter and a gesture-controlled virtual mouse.
