Summary
Eric Qasemi is a Senior AI Scientist and USC PhD candidate with eight years of experience building and deploying advanced NLP and multimodal ML systems. He has led research and engineering efforts across academia and industry, from instituting a novel preconditioned inference task for evaluating affordance reasoning in LLMs to shipping high-performance multimodal RAG ingestion pipelines and speculative-decoding LLM accelerations at Oracle. His work blends weak supervision, meta-learning, and neuro-symbolic techniques to improve commonsense and visual-language reasoning, and he has managed and mentored teams that secured NSF REU awards. Eric’s background spans embedded hardware and big-data research to end-to-end AutoML systems that outperform human baselines on D3M benchmarks, demonstrating an unusual depth across low-level systems and high-level model evaluation. Based in California, he combines publication-driven research with production impact, including a pending patent for inference acceleration. He’s driven by making ML models more interpretable and operational at scale, especially in multimodal and constrained-data settings.
8 years of coding experience
12 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at University of Southern California
Master's degree Computer Science, Master's degree Computer Science at University of Wisconsin-Madison
Bachelor of Engineering (BEng) Electrical and Electronics Engineering, Bachelor of Engineering (BEng) Electrical and Electronics Engineering at University of Tehran
English, Persian, Kurdish, Turkish