Aryan Sharma
@aryan.sharma
AI Intern at Bhashini
New Delhi, India
Aryan Sharma is an AI Intern with experience in optimizing machine learning pipelines. He engineered size reductions using 8-bit quantization and integrated ASR models into Triton inference servers. His academic background includes studies in IT & MI, and he has practical project experience in privacy-preserving ML and advanced network analysis.
Experience
AI Intern
Bhashini
Engineered a pipeline to compress torchscript modules using 8-bit integer quantization, resulting in a size reduction of 73% to 143MB. Developed a demo Android application for offline inference, leveraging the optimized ASR model for real-time processing. Integrated ASR and Intent and Entity Recognition models into a Triton inference server pipeline by converting models from ONNX format to TensorRT.
Education
University of Delhi (Cluster Innovation Centre)
B.Tech.
IT & MI