AI/ML Engineer & Co-Founder, OpenTrain AI (2022–Present)
Text
Fine-tuningEngineered data acquisition, annotation pipelines, and quality assurance workflows to produce large-scale training datasets for LLM and domain-adaptive fine-tuning. Led end-to-end ML pipeline development for training, fine-tuning, and serving models in production settings. Implemented distributed training and optimized data loading strategies to accelerate model training while maintaining dataset quality. • Oversaw annotation pipeline operations and dataset QA processes for 10M+ labeled samples. • Built and maintained MLOps tooling to evaluate experiments and monitor model performance. • Designed custom attention mechanisms for transformer-based domain tasks to improve accuracy. • Integrated AI capabilities into products using evaluated model outputs to improve engagement metrics.