You've been brought in as a Principal Data Scientist to a new organization where the data infrastructure is fragmented, documentation is minimal, and there's no clear data strategy. How would you, in your first 90 days, assess the current state, identify the most critical data-related problems impacting business objectives, and begin to lay the groundwork for a robust, scalable, and strategically aligned data science ecosystem, demonstrating immediate value despite the inherent ambiguity?

MECE Framework: 1. Assess (Weeks 1-4): Conduct stakeholder interviews (business, engineering, product) to map objectives, existing data sources, and pain points. Perform data infrastructure audit (tools, pipelines, quality, access). Review existing models/reports. 2. Prioritize (Weeks 5-8): Identify critical business problems addressable by data science. Use RICE scoring (Reach, Impact, Confidence, Effort) to rank projects. Focus on high-impact, low-effort wins. 3. Strategize & Execute (Weeks 9-12): Develop a phased data strategy roadmap (short-term wins, long-term infrastructure). Initiate a pilot project demonstrating immediate value (e.g., predictive model for a key metric). Establish initial data governance principles and documentation standards.

My first 90 days would follow a structured, iterative approach, leveraging the MECE and RICE frameworks. Initially, I'd conduct comprehensive stakeholder interviews (business, engineering, product) to understand objectives, current data usage, and critical pain points. Concurrently, I'd perform a rapid audit of the existing data infrastructure, identifying key data sources, pipelines, quality issues, and access limitations. This assessment phase (Weeks 1-4) would culminate in a prioritized list of data-related problems impacting business objectives, ranked using RICE scoring to identify high-impact, feasible projects.

Weeks 5-8 would focus on deep-diving into the top-ranked problem, developing a clear problem statement, and identifying available data. I'd then initiate a quick-win pilot project, aiming to deliver tangible, measurable value within the 90-day window. This could involve building a simple predictive model or an insightful dashboard addressing a critical business question. Simultaneously, I'd begin documenting key data assets and establishing initial data governance principles. By Week 12, I'd present the pilot project's results, a preliminary data strategy roadmap, and a prioritized backlog of future data science initiatives, demonstrating immediate value and laying the groundwork for a scalable ecosystem.