SOPHIA SASSOLI

• Engineered a global data architecture for a major motorcycle manufacturer - reconciled and cleansed 20M+ transactional records with a Snowflake-based medallion pipeline, driving material improvements in data integrity and reporting and identifying over $15M in cost savings. The supply chain model assessed scenarios across a global network of distribution centers, manufacturers, vendors, and dealers spanning motorcycles, parts & accessories, and apparel.
• Developed and validated a North American network model for a logistics service provider using SQL and Optilogic's Cosmic Frog for seven regional scenarios, resulting in a 5–10% transportation cost reduction and a 1.5% improvement in overall operating costs.
• Directed cross-functional delivery by leading weekly client sessions and coordinating workstreams with international teams in India, ensuring seamless project execution across global time zones.
• Mitigated significant project risk and enabled the realization of downstream storage analyses for a major industrial distributor by developing an XGBoost ML model to impute missing product dimensions for 5,700+ SKUs, achieving up to 91% confidence interval coverage and unblocking a stalled distribution center assessment.

• Standardized firm-wide methodologies by developing a reusable Python geocoding and distance script and a standardized Excel capacity model adopted as templates to accelerate timelines for subsequent engagements.
• Optimized facility capacity for two major assessments by modeling storage and automation alternatives, contributing to capital estimates and slotting plans to support future growth for a consumer-packaged goods manufacturer and distributor.
• Mentored junior analysts on data-driven methodologies, fostering a culture of technical excellence and structured thinking within engagement teams.
• Designed interactive decision-support tools using Tableau and Power BI to provide clients with scenario-based insights into fulfillment operations and capital investments.

• Built a software license compliance dashboard in Power BI using DAX and PowerQuery, surfacing inactive and unused licenses to drive cost reduction; also developed a Windows endpoint health dashboard for security and licensing audits across the enterprise.
• Performed data acquisition and manipulation in SQL across SCCM and ServiceNow data sources to support endpoint management and infrastructure reporting.

Developed a two-stage Python ML pipeline utilizing XGBoost and Random Forest with Bayesian averaging to predict game spreads with 64.3% win/loss accuracy. Leveraged NLP-informed feature engineering to capture nuances in non-power-conference data. Project sponsor praised the team's creative methodologies as a key differentiator beyond standard algorithmic approaches.

Projected $420K in annual labor savings and a 17.6% improvement in rack utilization for a Southeast Distribution Hub. Achieved results through a full-scale Simio simulation and the development of a custom VBA/Excel greedy optimization tool, presenting results in a 30+ page technical report and client presentation to faculty and industry advisors.