LLM AI Trainer (RLHF Evaluator & Prompt Engineer)
As a Senior Software Engineer, I participated directly in Large Language Model (LLM) fine-tuning and evaluation tasks. My responsibilities included reviewing LLM outputs, providing corrections, refining prompts, and curating datasets to improve response quality. I played a key role in aligning model responses with business requirements, using reinforcement learning from human feedback (RLHF) and prompt engineering. • Evaluated LLM responses for relevance, coherence, and accuracy based on pre-defined metrics. • Trained AI models by reviewing outputs and rewriting model answers to enhance clarity and correctness. • Created and refined prompts and datasets to help the AI understand intent and context more accurately. • Leveraged internal/proprietary tools for prompt + response writing, RLHF feedback, and dataset enhancements.