Mehdia Fatima

• Drove 50% growth in global pricing intelligence coverage, enabling strategic benchmarking across top hotel chains
(Hyatt, Marriott, Radisson) by aligning vendor pipelines and ingestion frameworks.
• Partnered with PMs and Market Managers to unlock $50M+ in incremental revenue, designing analytics on Beat
Meet-Lose (BML) metrics that shaped executive pricing strategy and partner pitches.
• Built and automated a real-time operations intelligence dashboard (SQL + Power BI) to track ingestion health, feed
drop rates, and hotel chain coverage—became a standard across global ops and engineering teams.
• Designed a team-wide performance scorecard that unified metrics across multiple functions and automated
conditional formatting based on metric thresholds—streamlined reporting and improved stakeholder visibility.
• Developed reporting pipelines using demand and clickstream data to support Market Managers in hotel partner
conversations—enabled more informed pricing negotiations and product feedback loops.

• Built a strategic geo-optimization model (Haversine + Excel Solver) to streamline warehouse placement—influencing
3 major client RFPs and reducing average fulfillment distance by 22%.
• Created a 7-module Power BI suite analyzing 1M+ rows of logistics data, helping 30+ enterprise clients achieve 15%
lower labor costs and 18% faster SKU retrieval through data-led decision-making.

• Improved ML model recall from 85% to 100% for Discover commercial content by designing advanced prompt
frameworks—automating operations and eliminating $300K/year in manual workload.
• Led content fairness audits uncovering gender and publisher bias in ranking models; findings led to 8 engineering
deep dives and algorithmic policy updates.
• Conducted A/B-like impact assessments of feed quality interventions across 3 languages—reduced low-quality
content by ~5%, improving user experience and policy compliance.
• Built real-time performance dashboards for 20+ global teams, improving vendor SLA visibility and cutting issue
resolution time by 22%.
• Implemented anomaly detection logic via SQL, reducing time-to-alert by 40% for critical issue buckets.