Medal enables millions of gamers to capture and share their best gaming moments and create memories together. Medal's 1M+ daily active users create 3M+ videos per day on our desktop and mobile applications. Your work will have a real impact on millions of people around the world! This role will be responsible for embedding Medal's database of billions of gaming clips to be contextualized more easily in the product, in addition to augmenting the embeddings via automated annotations and computer vision to create rich context around the clips. From a product perspective, this work will be used to improve library and social search functionality, and inform the ML recommendation system in Medal's social feed. In the future, the work will be used to improve the quality of the video editor via automation.
What we're looking for:
Depth of XP: 4+ years of video machine-learning, with a focus on image embedding and computer vision. Bonus for applicants who have XP in generative AI, familiarity with high throughput video inference/feature extraction
In-person: Looking to hire in NYC to contribute alongside the product team. 3+ days in the office
Ownership & attention to detail: You see things through, and can be responsible for end-to-end quality of complex features.
Code Performance: Excellent understanding of code performance and performance implications in production
Be result-driven: Everything we do is driven by the metrics and performance of the feature. We move fast and ship regularly
Measuring with metrics: Comfort with A/B testing and measuring results with usage metrics
Gaming: A passion for games and the gaming communities, and a user of Discord and other gaming-adjacent products. XP working on gaming-related projects is a plus
Our stack:
ML technologies: GCP, Bigquery, OpenCV, PyTorch, KubeFlow Pipelines, Airflow
Desktop Front-end: Electron, React, Redux, & other modern web-based technologies
