I led a project for Warner Bros in collaboration with ACME to develop a multimodal ML model that streamlines content search and editing. This model improves with each query, enabling detailed searches across a vast content library.

• Partnership: Secured collaboration with Warner Bros and ACME.

• Model Definition: Defined the multimodal ML model parameters.

• Data Management: Oversaw data collection, training, and validation.

• Contractual Obligations: Managed all contractual aspects to ensure clarity and compliance.

I developed an innovative model for automatic tracking of individuals on stage, simplifying video conversion from landscape to portrait for comedians and content creators.

• Scope Definition: Defined the scope of the tracking feature.

• Customer Discovery: Led customer discovery efforts to iterate and refine the feature.

• Quality Assurance: Led the efforts on QA to ensure the feature's reliability and performance.

I led the development of a predicted engagement model to help content creators convert long-form content into engaging short-form clips, identifying the most engaging moments automatically.

• Model Definition: Defined the parameters and functionality of the engagement model.

• Training Data: Utilized a dataset of a million hours of YouTube videos to train the model.

• Data Management: Oversaw the collection, training, and validation of datasets.

• User Experience: Ensured the system provided easy processing for users to receive clips based on engagement levels.

• Quality Assurance: Led QA efforts to ensure the model's accuracy and reliability.

• Project Management: Coordinated all aspects of the project to ensure successful delivery and implementation.

I led the initiative to host our own LLM models, creating ML models that automatically generate YouTube chapters for content creators, simplifying video organization and enhancing viewer engagement.

• Model Definition: Defined the functionality and parameters of the chapter generation model.

• Data Management: Oversaw data collection, training, and validation of datasets.

• User Experience: Ensured the tool effectively simplified content organization for creators.

• Quality Assurance: Led QA efforts to maintain the model's performance and reliability.

• Project Coordination: Managed all aspects of the project to ensure successful delivery and implementation.

I spearheaded the development of an AI model that transcribes audio into text with 80% accuracy, streamlining the creation of video captions and enhancing accessibility and viewer engagement.

• Model Definition: Defined the parameters and functionality of the transcription model.

• Data Management: Oversaw data collection, training, and validation of datasets.

• User Experience: Ensured the model provided a simple and effective way for creators to generate captions.

• Quality Assurance: Led QA efforts to ensure transcription accuracy and reliability.

• Project Coordination: Managed all aspects of the project to ensure successful delivery and implementation.

I spearheaded the development of advanced AI models for Universal Tennis, revolutionizing tennis analytics. My responsibilities included securing the partnership, defining AI models, overseeing data collection, and managing training and validation datasets, along with handling all contractual obligations. The developed models included:

• Movement Recognition: Analyzing different types of swings.

• Biomechanics: Calculating angles and body positions for performance insights.

• Object Tracking: Monitoring ball trajectory and speed.

• Court Mapping: Creating a 2D model of the court to visualize ball landing spots.

I led the initiative to develop AI models in combat sports, collaborating with Paradigm Sports and local venues to create advanced movement recognition models that predict real-time movements. My responsibilities included securing partnerships, defining models, overseeing data collection, and managing training and validation datasets, along with handling all contractual obligations. The developed models included:

• Movement Recognition: Predicting and analyzing real-time movements during combat sports events.

• Athlete Performance: Providing insights into athlete movements to enhance training and performance.

• Real-Time Analysis: Enabling coaches and athletes to receive immediate feedback during practice and competitions.

• Data Visualization: Creating visual representations of movement patterns to aid in strategy development.

Collaborating with popular softball coach Morgan, I spearheaded the development of an AI model designed to differentiate between multiple athletes using face recognition technology. This model aims to auto-clip long-form videos based on athlete detection, streamlining video analysis and review processes. My responsibilities included securing the partnership, defining the model, overseeing data collection, and managing training and validation datasets, along with handling all contractual obligations. The developed models included:

• Face Recognition: Differentiating between multiple athletes using advanced face recognition technology.

• Auto-Clipping: Automatically clipping long-form videos based on athlete detection.

• Streamlined Video Analysis: Enhancing video review processes by providing precise clips for analysis.

• Data Management: Overseeing the collection, training, and validation of datasets.

Partnering with renowned QB coach Jordan Palmer, I led the development of a machine learning model to perform advanced biomechanics analysis on athletes. This model analyzes and optimizes athletes' movements, providing critical insights for improving performance. My responsibilities included securing the partnership, defining the model, overseeing data collection, and managing the training and validation datasets, along with handling all contractual obligations. The developed models included:

• Biomechanics Analysis: Performing advanced analysis of athletes' movements.

• Performance Optimization: Providing critical insights to optimize athlete performance.

• Movement Analysis: Analyzing body mechanics to improve technique and efficiency.

• Data Management: Overseeing the collection, training, and validation of datasets.

In collaboration with Paradigm Sports and local venues, I led the development of sophisticated AI models designed to track and predict movements between two athletes during sparring sessions. These models provide real-time insights, enhancing training and performance analysis in combat sports. My role included securing partnerships, defining the AI models, overseeing data collection, and managing training and validation datasets, along with handling all contractual obligations. The developed models included:

• Movement Tracking: Monitoring and predicting movements between two athletes during sparring sessions.

• Real-Time Insights: Providing immediate feedback to enhance training and performance analysis.

• Performance Analysis: Delivering detailed analytics to improve athletes' techniques and strategies.

• Data Management: Overseeing the collection, training, and validation of datasets.