- Computer Vision
- Model Deployment
- Natural Language Processing
- Predictive Analysis/Time-Series
- Amazon Web Services
James Conley studied computer science and mathematics at Connecticut College. He twice presented in the ML/AI tracks of the MIT URTC on the creation of GPU accelerated algorithms with NVIDIA CUDA technology, LSTM-based game controllers, and Xpilot-AI, a library for developing Xpilot bots. Prior to joining SFL Scientific, he worked at Travelers Insurance Innovation Center, taking part in each stage of development for CNN-based models for image classification and segmentation tasks. He developed interfaces for and oversaw mass annotation to develop internal datasets, produced models leveraging Tensorflow/Keras, and deployed models to cloud-based infrastructure.
Since 2019, James works as a data science consultant at SFL Scientific to develop and deploy models solving problems across AI domains, including projects centered on business analytics (regression), natural language processing (clustering, NER, language models), and computer vision (classification, object detection, semantic segmentation). James is an AWS Certified Developer and specializes in a broad range of languages and libraries including C/C++, R, ReactJS, Tensorflow, Gensim, Pandas/Numpy/Sckit-Learn, Docker, Flask, and multiple cloud platforms.