Isaiah
De La Rosa

Fluffy is an all-in-one internal analytics platform I built and deployed for Feeding San Diego. The organization needed a way to surface distribution trends, fulfillment KPIs, and ad-hoc reporting without relying on manual exports — so I built it from scratch, deployed it to the cloud, and handed off a tool the team could actually use.

• Deployed a Python web app on Google Cloud Run with KPI dashboards, automated reports, and user authentication
• Engineered a pandas + Plotly pipeline to analyze order trends, fulfillment activity, and period-over-period variance
• Built a natural language search tool with custom tokenization to retrieve distribution and reporting metrics
• Modeled throughput, inventory cycle times, and efficiency to forecast capacity needs for warehouse expansion

Lumnus is a student-run for-profit consulting group at UCSD. With limited consultant bandwidth and a growing project pipeline, the team needed a systematic way to evaluate which engagements to take on. My team built a weighted decision matrix that quantified tradeoffs across eight dimensions so leadership could make consistent, defensible capacity decisions.

• Designed a multi-variable scoring matrix weighing client value, consultant growth, pay, effort, time, cost, complexity, and strategic fit
• Built configurable weightings so leadership could adjust priorities based on organizational goals each semester
• Implemented a feedback and analysis framework to capture post-project learnings and refine future scoring
• Applied the model to a competitive landscape assessment for NanoMood's remote Ketamine monitoring platform

When I took office as Chapter Treasuer (and later Chapter President) I inherited a $5,000 deficit with no financial documentation or formal budgeting process. I built a structured Excel model to stabilize the chapter's finances, map a debt recovery path, and create a reusable framework for future treasurers.

• Modeled a quarterly budget with built-in reimbursement workflows and variance tracking against actuals
• Forecasted multi-quarter recovery path to resolve $5,000+ in inherited deficits while maintaining operations
• Standardized reimbursement documentation to reduce treasurer onboarding time and processing errors
• Built growth goal scenarios to guide chapter expansion planning and long-term financial stability

Purchase order creation at Hologic was slow and error-prone due to fragmented supplier data across Oracle ERP, Agile PLM, and manual spreadsheets. I built a centralized supplier-item database linking materials, vendors, and catalog numbers to support automated MRP workflows and significantly cut PO creation time.

• Built supplier-item database linking 1,500+ materials, vendors, and catalog numbers across procurement systems
• Validated Oracle Agile PLM master data for 1,500+ items against purchasing, compliance, and receiving requirements
• Redesigned risk assessment Excel calculator to identify production risks and support supply continuity planning
• Standardized buyer documentation to support automated MRP purchase order workflows