Awni Hannun
Mountain View, California, United States
Summary
🤩
Rockstar🎓
Top SchoolAwni Hannun is a research scientist with 13 years of experience building high-impact machine learning systems across industry leaders including Apple, Zoom, Facebook, and Baidu. He co-created the MLX framework at Apple and has a track record of making language and speech models more compute- and sample-efficient. At Baidu he led the team behind Deep Speech, a widely recognized breakthrough in end-to-end speech recognition, and later helped found Zoom’s AI lab to advance meeting intelligence and speaker analytics. His open-source work includes production-ready deep learning projects for ECG arrhythmia detection and end-to-end speech-to-text, demonstrating breadth from healthcare to audio. Trained in mathematics, economics, and computer science (PhD from Stanford), he blends rigorous theory with production engineering to ship robust ML systems. Peers describe him as a researcher who consistently translates novel ideas into scalable, real-world deployments.
13 years of coding experience
10 years of employment as a software developer
BA, Mathematics, Economics, BA, Mathematics, Economics at Dartmouth CollegeDoctor of Philosophy (PhD), Computer Science, Doctor of Philosophy (PhD), Computer Science at Stanford University
Github Skills (14)
data-preprocessing10
keras10
pytorch10
machine-learning10
speech-recognition10
rnn-model10
deep-learning10
tensorflow10
n10
python10
data-extraction9
data-loading9
scikit8
scikit-learn8
Programming languages (16)
Github contributions (5)
awni/speech
Jun 2017 - Apr 2021
A PyTorch Implementation of End-to-End Models for Speech-to-Text
Role in this project:
Back-end Developer
Back-end DeveloperContributions:151 commits, 7 PRs, 56 pushes in 3 years 11 months
Contributions summary:Awni primarily contributed to the speech-to-text project by implementing core features, specifically a Librispeech dataset downloader. They set up the early model structure and also included a preprocessing script to prepare the data. Furthermore, they added utilities for converting audio files and included a training script which indicates they are involved in the model training process.
pytorchspeech-to-text
awni/ecg
Oct 2016 - Apr 2022
Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network
Role in this project:Data Scientist
Data ScientistContributions:126 commits, 1 PR, 18 pushes in 5 years 6 months
Contributions summary:Awni appears to be a data scientist working on the development of a deep learning model for arrhythmia detection and classification within the context of ambulatory electrocardiograms. Their contributions involve setting up and integrating a WFDB library for processing ECG data, creating scripts for downloading and extracting data from the MIT-BIH Arrhythmia Database. Moreover, the user designed, built, and began training a neural network (RNN) model for ECG analysis, including defining model architectures and a training pipeline.
classificationneural-network
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.