Teams at SumerSports

Lead design, training, evaluation, and production deployment of deep-learning models for player-tracking. Drive flagship modeling threads from problem definition to production, collaborate with football experts on features and labels, mentor mid-level engineers, and build experimentation infrastructure to increase research throughput.

Design, build, and maintain reliable data pipelines and lakehouse assets for batch and streaming workloads. Implement ETL/ELT in Python and SQL, orchestration and monitoring (Airflow/Kubernetes), and retrieval/RAG/vector search pipelines for LLM and AI applications. Ensure data quality, lineage, versioning (Delta/Parquet/Iceberg), cost/performance optimization, and collaborate with MLOps and sports data teams.

Design, train, and evaluate computer vision models for sports video (detection, tracking, pose, event recognition, re-ID). Run experiments, error analysis, dataset and evaluation design, and collaborate to productionize models with attention to reproducibility and efficiency.