Big Data & Analytics Summit - Data Science Challenges @ Lazada
Eugene Yan
Lazada's data science team grew from 4-5 people to 40 over three years while building machine learning systems to replace manual business processes. Key challenges included balancing automated algorithms with business override requirements, managing development speed versus production stability, and prioritizing short-term business needs against long-term innovation. The team addressed these by conducting A/B tests to quantify the impact of manual interventions, investing in code quality and documentation after initial rapid development, and using timeboxed skunkworks projects to balance exploration with business impact.
Why it matters
Technical challenges easy compared to business and people issues. Sharing at the BDA Summit.
