• Full stack web application development to facilitate the genetics lab operations using C# and SQL.

Created full stack .NET application to allow lab directors to modify genetic mutation tables, streamlining lab workflows and operations.

• Wrote Python utilities to protect patient PHI (Protected Health Information) compliant w/ HIPAA safe harbor standards.

• Automation of NGS (Next-generation genetic sequencing) analysis using Python.

• Own my design and coding decisions, work with the lab users to get feedback and take responsibility for bug fixes, improvements, etc.

• Technologies used: C#, .NET Core, SQL, Python, JavaScript

InfoPost is a startup using AI to detect fake or misleading news and determine trust scores for news articles. My responsibilities included:

• managing a small team of NLP interns to contribute directly to production software (sentiment analysis, bias detection, web scraping)

• End to end ownership (training, testing, optimizing and deploying) of large scale language models to detect misleading headlines.

• Creating rich data sets for training and validating Transformer models using Prodigy and Python scripts.

• Deploying models to GCP production systems processing over 1000 articles per day over last 6+ months.

• Maintenance (code review, refactoring, integration) of AI backend (10k+ lines of code).

• Technologies Used: Python, PyTorch, TensorFlow, transformer models (BERT, XLNet, etc.), Google Cloud Platform (GCP), Spacy, NLTK, gensim, Flair.

Analytics and Machine Learning/Computer Vision R&D. Worked closely with deep learning research scientists, creating datasets to optimize and train PyTorch models.

Worked in a team with undergrad and graduate students to create a system of lesson plans and teaching modules to teach high school level English using music & popular culture.