ASA

• LLM model development

App Ads Core ML

• developing ML Ranking Models

AI Infra

• Automated and deployed an ML pipeline for the ESG Data Analytics thermal team using python and Jenkins for Linear Regression and RNN models that achieved under 0.5 degrees celsius root mean squared error.

• Developed a novel solution for determining when models should be activated and deactivated to save CPU cycles through multivariable optimization.

• Performed feature engineering and trained Random Forests, Gradient Boosted Trees, and Neural Networks to classify whether cellular devices were using 5G with 0.75 F1 score on highly imbalanced datasets.

• Project Leader for Industry Consulting Project with FakeNetAI on deepfake audio detection.

• Devised new ResNet-LSTM models to classify audio segments from multiple datasets with an average of 0.9 F1 score for FakeNetAI’s audio deepfake detector.

• Deployed model on to a web server with AWS and streamlit backend.

• Built a pipeline that analyzes the content and semantics of sales calls in real time to automate CoPilot's cue card system for sales representatives.

• Created custom unsupervised PyTorch word models with importance weighting and embeddings to achieve an F1 score of .75 trained on unlabeled data.

• Analyzed representational similarity divergence of various vision models in hierarchical pretraining strategies under Prof. Kurt Kuetzer.

• Published Self-Supervised Pretraining Improves Self-Supervised Pretraining to the WACV 2022 conference.