Suleman Ibrahim

AI LLM Trainer

Turing • Freiberufler

Dec 2023 - Present • 2 yrs 5 mos

Training and fine-tuning large language models by crafting high-quality prompts and systematically evaluating model outputs across diverse domains. Collaborating with global AI teams to enhance model reasoning, factual accuracy, and instruction-following capabilities while developing domain-specific datasets and test scenarios, particularly for software development and technical fields. Continuously analyzing model behavior and providing structured feedback loops that are contributing to a measurable 20% improvement in response quality. Writing, reviewing, and refining thousands of structured prompts to support supervised learning and reinforcement learning workflows, identifying edge cases, and helping mitigate hallucinations and inconsistencies in generative AI outputs. Actively participating in ongoing research initiatives focused on optimizing prompt formats and instruction tuning for reliable, commercial-grade AI systems.