Parsa Farshadfar

Led weekly tutorials for the Design and Analysis of Algorithms course, breaking down complex concepts into practical problem-solving strategies to boost student comprehension. Orchestrated exam review sessions, guiding peer-to-peer technical discussions and fostering a collaborative problem-solving environment.

Led end-to-end delivery of production ML models, defining technical direction from problem framing and requirements through deployment, monitoring, and iteration. Owned the model development lifecycle operating model, establishing clear stage gates, artifacts, and handoffs to accelerate delivery while improving consistency. Partnered with Risk, Compliance, and Technology stakeholders to ensure audit-ready solutions by design, aligning documentation, testing evidence, and approvals with governance expectations. Set roadmap and priorities for a portfolio of ML deliverables, balancing business impact, engineering effort, and operational risk. Standardized reusable components to reduce rework and increase adoption. Drove quality and reliability practices across ML development including data profiling, data drift, performance monitoring, explainability, bias/fairness, reproducibility, versioning, and production monitoring. Provided technical leadership and mentorship to scientists and engin

Built repeatable data pipelines and feature datasets, applying data quality checks and documentation to improve reliability and reuse. Built and maintained AI/ML models (PyTorch, TensorFlow, XGBoost, LLMs, Scikit-learn) in clean modular code with efficient version control. Monitored post-release performance using defined KPIs and feedback loops, triaging issues and recommending retraining or corrective actions based on pre-approved action plans. Produced clear report artifacts to support governance reviews and future maintenance. Improved team efficiency by contributing reusable code, templates, and internal knowledge transfer meetings (how-tos, runbooks, and workshops).