01 / VSCO · 2024–Present
ML-Driven Creative Tools & Feed Recommendation
7%
Engagement & retention lift
5%
Trial conversion
0→1
First ML feed rec system at VSCO
The Challenge
VSCO had rich creative tools but no intelligent surface to help users discover content or editing inspiration. The app needed both a smarter feed and AI-native editing capabilities to stay competitive in a rapidly shifting creative landscape.
What I Did
I defined the AI/ML product vision end-to-end — owning model selection, integration constraints, evaluation criteria, and rollout strategy for a suite of ML-driven creative editing tools built on external models. In parallel, I led 0→1 delivery of VSCO's first ML feed recommendation system on iOS, partnering directly with ML engineers to define training data, offline evaluation metrics, and online A/B tests.
Critically, I balanced model quality, latency, and inference cost throughout — and made targeted contributions to the iOS codebase to prototype changes before committing engineering resources.
Beyond the Feature
I built and now teach VSCO's experimentation and analytics framework, mentoring PMs and engineers across teams to institutionalize data-driven decision making. I also led vendor negotiations and data platform modernization efforts, and partnered with leadership to define VSCO's AI ethics standards and model evaluation best practices.