Kushal Dulla

Background

About Kushal Dulla

I've always been fascinated by the different ways in which technology and humans can interact and through my various research opportunities, I've been able to explore these interactions through several projects in the field of Natural Language Processing, Machine Learning, and Large Language Models. With a strong foundation in programming languages such as Python, C++, and Java, coupled with expertise in various machine learning frameworks including PyTorch and TensorFlow. I am currently an Engineer with the Cloud ML team at Qualcomm, where I work on onboarding and optimizing Large Language Models, Automatic Speech Recognition models, and any other ML models required by customers to specialized hardware. I have further experience working in industry-focused research through my role as an ASR Research Developer Intern at Nuance Communications, where I researched and evaluated the utilization of LLMs as a speech recognition correction tool. In addition to industry experience, I was heavily involved in research at Johns Hopkins Whiting School of Engineering, Center for Language and Speech Processing, where I developed a multilingual fact verification dataset, based on FEVER, an existing monolingual dataset, and using cross-lingual Wikipedia articles. At Carnegie Mellon University, I similarly worked on NLP-focused research projects that gave me an understanding of several different aspects of the field, including speech recognition, conversational AI, and vision-and-language-based navigation. I thrive in innovative environments where I can advance the state-of-the-art in NLP and Machine Learning and am eager to apply my skills and knowledge to real-world problems and challenges.