AI Engineer Intern

Responsibilities & Contributions

• Built a social media simulation system for CritterCircle, Crittero's social platform for pet owners and animal rescues, generating 100,000 labeled posts across seven persona types with seeded random generation and time-series activity modeling (1,700+ lines of TypeScript).
• Developed a TensorFlow/Keras recommendation model achieving 80% accuracy — a 30-percentage-point gain over a random-posts baseline — with persona-type, content-category, and temporal embeddings, plus a collaborative-filtering cold-start fallback for new users.
• Created a configurable CLI testing framework in Python (700+ lines) with performance metrics and reproducibility validation, enabling systematic, repeatable evaluation of recommendation algorithms.