Over the past decade, I've worked in real-time logistics across Latin America-first with Cabify, a ride-hailing platform similar to Uber that focuses on corporate and private transportation in regulated markets, and later with Rappi, one of the region's largest "superapps," which handles on-demand deliveries for food, groceries, pharmacies, cash, and even financial services.
In these environments, we faced intense volatility: unexpected demand surges, political protests, unreliable infrastructure, and rapidly shifting user behavior. At first, we made the classic mistake-jumping into modeling without solid data pipelines. But we quickly learned that without real-time, clean, validated data, even the most elegant models fail under pressure.
By shifting our focus toward data quality first, we were able to implement predictive AI systems that matched delivery drivers with peak demand zones, improving operational efficiency and customer satisfaction. The most impactful result came from automating our payout validation process using Python scripts and open banking APIs-cutting error rates in weekly courier payments from over 40% to under 5%.
These weren't theoretical models. These were decisions that affected the daily income of thousands of drivers across Latin America.
I'd like to hear from others who are applying AI and decision science in complex, high-friction environments:
How do you prioritize data integrity before modeling?
Have you faced ethical or operational challenges when optimizing at scale?
Is anyone working on predictive systems for real-time logistics or delivery operations under infrastructure constraints?
I'm currently writing a book that connects AI, cybersecurity, and last-mile logistics, using Latin American case studies to explore scalable, ethical and practical solutions. I'd love to exchange ideas or collaborate with others building bridges between operations research and the messy, urgent reality of emerging markets.
------------------------------
Dámaso Diaz
Systems Engineer | Logistics and Artificial Intelligence Specialist
Orlando, Florida
------------------------------