Staff Data Analyst

Role Overview: 

Grailed is looking for a Staff Data Analyst / Data Scientist to drive personalization, recommendation, and product marketplace improvement efforts. 

The ideal candidate is able to think like a Grailed user as well as a business owner – understanding how data impacts a fashion-forward user experience and also how it is generated and leveraged – while bringing a strong technical background to the role.

Specifically, the role requires an understanding of dimension reduction techniques, predictive modeling (statistical or ML), and other advanced analytic methods.

This key role will operate at the intersection of Data, Product, Engineering, and Marketing, working crossfunctionally to develop compelling data products to help buyers find their Grail and sellers maximize their GMV.

This role will work with our data in Snowflake, develop models in Python, collaborate with ML engineers to structure data for consumption, and coordinate with Product and business unit leaders to align data product development with business objectives. 

Responsibilities:

• Must be able to form a high-level perspective on objectives across departments in the organization and how advanced data methods might solve complex business problems
• An ideal candidate is able to autonomously and proactively identify business problems that could benefit from data solutions, whether it be application of existing models or the need for the development of new model(s), and take ideas through all phases, from proposal to alignment to execution

• Establish best practices for training and development of data models; performance evaluation and monitoring; maintenance, etc
• Own the deployment of trained models into production in collaboration with Data or ML Engineers. You will be responsible for ensuring reliable, observable deployment into Snowflake using DBT, integrating with existing data pipelines and platform infrastructure, and maintaining version control of code and configurations via Git.

• Manage portfolio of existing Grailed Data Zone products, which currently includes the following domain areas (with the expectation to expand):
  • Similarity and Personalization for Recommendations: use of historical data to quantify similarity or dissimilarity of marketplace objects such as buyers, designers, sellers, collections, etc., empowering Grailed to segment and rank order product experiences such as the Grailed home page; recommendations on listing pages and designer pages; targeted email and push campaigns, etc
  • Listing Quality Assessment: as a marketplace with limited controls on new inventory, we use advanced data methods to evaluate and segment newly published listings based on their attributes and the historical outcomes associated with those attributes. Applications include search optimization; marketplace health; trend analysis; valuation (e.g. comparing one power seller to another based on expected outcomes of their active inventory). etc
  • For the right candidate, additional, but less frequent responsibilities may arise with respect to performance marketing initiatives – such as geo targeted incremental lift studies.

Additional areas of existing responsibility:

• Evaluate model performance and iterate to improve accuracy and effectiveness. This includes using A/B testing to validate the impact of personalization initiatives and communicating results to stakeholders.
• Mine user data to identify opportunities for personalization improvements. This includes defining and tracking KPIs related to personalization effectiveness.
• Develop and maintain data models in Snowflake to support analytical and reporting needs, providing insights to business stakeholders across various departments.
• Use Python to create ML models and structure the resulting data into a consumable flow.
• Develop user-to-user mapping capabilities using graph databases and vector embeddings to enhance personalization.
• Utilize search technologies (i.e. Algolia, AWS OpenSearch) to enhance product discovery and personalization.
• Analyze message content to detect potentially fraudulent activities, such as identifying keywords or phrases associated with scams, requests for off-platform transactions, or attempts to phish for personal information.
• Collaborate with product managers, engineers, designers, and business stakeholders to understand their data needs and provide data-driven solutions.

Requirements:

• Education
• Graduate degree in data science, analytics, mathematics, machine learning, computer science, or related field a plus
• Demonstrated track record of applying analytical skills in a product or business setting may substitute for formal advanced education.
• Professional Experience:
• 8+ years of relevant work experience in a data or quantitative role, demonstrated success in a startup, high-growth or faced paced organization
• Experience in marketplace, e-commerce, or fashion/retail domains preferred
• Experience with web + App product environment preferred
• Experience with Marketing analytics a bonus
• Demonstrated success in nontechnical, crossfunctional partner communication
• Ability to tell a story with data, explaining complex concepts or results to audiences ranging from C-suite to IC levels
• History of mentoring or developing teammates
• Ongoing learning (e.g. relevant certifications; open-source contributions; personal projects; etc.) is a plus. 
• Technical Competencies
• SQL (Expert)
• Python (Expert)
• Familiarity with Looker preferred
• Familiarity with Amplitude preferred
• Familiarity with DBT preferred

• Math and Statistics
• Proven expertise in advanced statistical modeling, causal inference, experiment/test design, and working knowledge of machine learning algorithms.

• Data Science and Engineering 
• Expert level proficiency in Python for data manipulation, statistical analysis, and model development
• Practical experience with vector databases and embeddings for tasks like user-to-user or user-to-item mapping, semantic search, or item similarity preferred
• Experience with Snowflake for SQL and data-warehousing preferred
• Experience with DBT for building modular, version-controlled data transformations preferred
• Experience with Git for collaborative code development and review preferred

• Machine Learning and AI
• Experience in designing, developing, deploying and optimizing Personalization and Recommendation products at scale
• Experience building models to assess item/listing quality (as defined by likelihood of sales), classify listings, and use NLP on unstructured text
• Experience modeling time-series forecasts for market trends, seasonality, demand prediction and other relevant KPIs