Susan Eraly
Data Scientist
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolSusan Eraly is a data scientist with 11 years of experience applying deep learning and NLP techniques to production-grade tooling and model import pipelines. She transitioned from ASIC and physical design engineering at NVIDIA and HP into machine learning engineering, contributing substantial fixes and examples to the Deeplearning4j ecosystem—most notably BERT inference and TensorFlow model import utilities used by JVM-based ML practitioners. As an independent contractor and former Skymind engineer, she blends low-level systems rigor with practical model deployment experience, improving usability of Keras/TensorFlow imports and NLP iterators. Based in New York with roots in the Bay Area, she brings a pragmatic, open-source-first approach to making complex models easier to run in enterprise Java environments.
11 years of coding experience
10 years of employment as a software developer
Bachelor of Engineering (BE) Electrical Engineering, Bachelor of Engineering (BE) Electrical Engineering at Stony Brook University
English, Malayalam, Hindi
Stackoverflow
339reputation
10kreached
20answers
0questions
Github Skills (21)
artificial-intelligence10
machine-learning10
java10
javas10
keras10
deep-learning10
tensorflow10
neural10
neural-network10
nlp10
bert9
hash6
dl4j6
csh6
vim6
Programming languages (6)
Github contributions (5)
deeplearning4j/deeplearning4j-examples
Jun 2016 - May 2020
Deeplearning4j Examples (DL4J, DL4J Spark, DataVec)
Role in this project:
ML Engineer
ML EngineerContributions:157 commits, 61 PRs, 133 pushes in 4 years
Contributions summary:Susan contributed to model import examples, specifically focusing on TensorFlow models within the Deeplearning4j framework. Their work included fixing paths for TensorFlow model imports, creating a BERT inference example for sentence pair classification, and adding scripts related to downloading models. The user also modified a Word2Vec sentiment example to include a model download if one doesn't exist.
deeplearning4jpythonzeppelin-notebookdeep-learningmachine-learning
deeplearning4j/deeplearning4j
Apr 2016 - Mar 2020
Suite of tools for deploying and training deep learning models using the JVM. Highlights include model import for keras, tensorflow, and onnx/pytorch, a modular and tiny c++ library for running math code and a java based math library on top of the core c++ library. Also includes samediff: a pytorch/tensorflow like library for running deep learn...
Role in this project:ML Engineer
ML EngineerContributions:583 commits, 106 PRs, 185 pushes in 4 years
Contributions summary:Susan primarily contributes to the deeplearning4j repository by modifying and improving Keras model import functionality and BERT iterator tests and utilities. They are fixing bugs, adding convenience methods, and cleaning up existing code related to model import and NLP tasks. The contributions demonstrate a focus on improving the usability and functionality of model importing and NLP capabilities within the DL4J framework.
deeplearning4jpythonautomatic-differentiationneural-netskeras-tensorflow
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.