At Turing, I build async FastAPI and PostgreSQL back ends for production-grade replicas of enterprise SaaS products used to train and evaluate AI agents. Each session has isolated, seeded database state and a run-scoped token, making sessions reproducible and parallelizable.
At CuddlyNest, I worked on the core team for its travel booking platform. I built search middleware that combined live hotel availability and pricing into a ranked results API, and supplier data pipelines using Kafka, RabbitMQ, and ETL jobs.
At Xeltec, I worked on BrickSkill and PaddleX, including a Python course-generation service and a FastAPI and Redis booking back end. I’ve also built LLM chatbots and RAG pipelines for client products, and my selected projects include a WebAR platform and a real-time voice AI SaaS for restaurants and hotels.

