Lead a custom AI dev shop and AI consultancy in Charleston SC. 9 successful projects in our first year, numerous repeat clients.

Senior developer working on LexisNexis Protege Developed custom RAG pipeline to ensure accurate answers to user questions based on legal database Built custom pdf and word doc extraction tools to detect minute formatting changings and group relevant document sections/sub-sections Designed custom tone and style extraction to detect and extract relevant writing style and tone features from user-supplied documents Designed rewriting agent to rewrite model outputs based on extracted styles and tones

Lead developer for a large multi-object tracking computer vision project - Multi-Object Tracking and Automated Injury Identification for Athletes Acted as project manager for teams of 3-6 engineers Developed custom Llama based LLM for extracting discrete fields from clinical notes Utilized Kanban boards, issue trackers, Gantt charts, architecture diagrams, and more for PM responsibilities Presented at Ohio AFP conference on potential usage/impact of AI in banking and finance Scoped around a dozen projects for prospective clients Responsible for setting and keeping to timelines Reviewed code from other contributors and provided feedback Founded a deep learning paper club to ensure myself and colleagues are up to date with latest research advances Implemented unit tests to ensure effective scalable code

Developed and implemented machine learning models to help detect fraud and overpayments Used clustering (intelligent vectorization and k-means) to develop member profiles Discovered widely outdated infrastructure and updated it to most recent versions (python, conda, cuda, etc.)

Designed and taught a PhD level course on applying machine learning and deep learning techniques to solve both business and clinical issues in the medical industry

Lead a multidisciplinary team of engineers, scientists, and doctors, to develop artificial intelligence (ML and DL) solutions for clinical and business problems Technical lead and PM on over a dozen projects in predictive analytics, image processing, and natural language processing – taking projects from raw idea to final product Responsible for all data science hiring, developed a custom hiring process Used convolutional neural networks (CNNs) for segmentation, object detection, image classification, etc. both in 2D and 3D applications Used NLP techniques include tokenization, TF-IDF encoding, vocabulary selection, LDA, etc. Installed and managed a local multi-GPU Linux workstation for cost savings and computing speedup Used classic ML techniques such as decision forest, SVM, logistic regression, linear regression, etc. Utilized data from EHR, imaging results, physician notes, provider efficiency data, patient messages, and more

Performed ML-based sentiment analysis on the business and its competitors utilizing Twitter data Automatically identified and analyzed abnormal social media events related to company interests

Worked within a team of PhD, masters, and undergraduate students to create pipelines to collect and preprocess social media data in real time for use in machine learning algorithms Utilized NLP on twitter data to predict outbreaks of food poisoning in real time, published results in AAAI 2016