siriki Soumahoro

● Architected MVP from scratch, balancing speed, cost, and long term scalability trade-offs ● Design a modular monolith Api architecture, optimized for scalability and future transition to microservices. ● Built a real-time recommender system, improving user engagement and retention through personalized product feed. ● Design sql database schema to support scalability and maintain data integrity. ● Implement a feed serving api for delivering personalized product recommendations with low latency. ● Developed a TikTok-style video streaming service optimized for performance on low bandwidth connections. ● Engineered an Uber-style delivery service, integrating real-time driver tracking. ● Built a comment service supporting nested replies, likes and moderation tools. ● Implemented a full text search engine, enabling fast relevant product discovery. ● Led a junior frontend team, accelerating feature development, improving app performance and mentoring team members for growth.

Developed and implemented a scalable deep learning model for KYC document validation, achieving a 10x reduction in enterprise costs. Contributed to constructing a text classifier for streamlining customer transaction categorization. Engineered a machine learning model for monthly batch inference, proactively pinpointing high-churn probability customers Executed statistical tests to evaluate the impact of a new feature on customer behavior. Developed a credit scoring model for customers with no prior credit history using fundamental algebra and a proxy target machine learning model.

Skills: Java