Squarespace

The Squarespace Machine Learning team is growing! We source projects from all over the company, ranging from direct integrations of NLP & computer vision into our products to behind-the-scenes services that improve efficiency in our customer help center. Our team prizes stakeholder-focused development of research and project deliverables while also taking time out for reading groups, conference attendance and Hack Week projects that pursue educational and research goals. The ML team uses a Python-focused development stack with all the usual ML and data processing libraries. We have a framework for rapidly deploying our models behind lightweight Flask web applications in a managed Kubernetes environment and to thrive in our team you should enjoy the full range of backend engineering tasks that go along with this. You won't be developing ML models in isolation, but instead demonstrating leadership and ownership over deployment strategies, architecture choices to solve cold-start problems, defining observability and monitoring strategies and working with stakeholders to define validation metrics that don't merely represent statistical considerations like overfitting or precision & recall but go further to develop holistic understanding of model success in a broader product development context. We are seeking a senior machine learning engineer with a highly developed skill set in software engineering, architectural design and experience mentoring junior engineers and collaborating with multiple stakeholders to deliver projects. You should be skilled in the application of machine learning both on the statistics side and the backend engineering side and comfortable moving between the two sets of responsibilities as the situation demands. We welcome applicants with experience in any variety of machine learning or applied statistics context (e.g. quant finance, time series analysis, audio processing, etc.), though the candidate should show an enthusiasm for information retrieval, search and recommendation, and computer vision models because these are at the forefront of what we do.