About the job
Job Purpose
Responsible for designing, building, and operating a unified machine learning infrastructure that standardizes the lifecycle of models from research to production. ridges the gap between batch-oriented data in Snowflake and real-time inference on Kubernetes (K8s), ensuring high-quality software engineering standards and reducing implementation redundancy across all data science squads.
Key Result Key Responsibilities
Designs and maintains the core ML Platform architecture, integrating Snowflake/Snowpark with Kubernetes for hybrid workload support.
Designs and delivers domain specific end to end data science products, including flight revenue forecasting, customer retention, ground operations, and others.
Automates end-to-end ML pipelines including data ingestion, model training, and continuous deployment using Apache Airflow and GitLab CI/GitHub Actions.
Builds and manages a centralized Feature Store and Model Registry within Snowflake to ensure consistency between training and serving.
Key Result Key Responsibilities-Continued
Implements comprehensive observability systems for monitoring model performance, data drift, and system health using Prometheus, Grafana, and Evidently AI.
Develops reusable FastAPI or gRPC service wrappers for real-time model serving.
Establishes "paved road" workflows (CI/CD, unit testing, modular code) for the broader data science team.
Qualifications (Academic, training, languages)
Bachelor's or Master's degree in Computer Science, Software Engineering, or a related field.
Fluent in English Language.
Deep expertise in machine learning, statistics, and applied modeling Mastery of Kubernetes, Docker, and Infrastructure as Code (Terraform).
Expert-level Python (OOP, modular design) and SQL.
Proficiency in serving models built with LightGBM, XGBoost, TensorFlow, and PyTorch.
Strong understanding of production ML systems and trade-offs Ability to influence architectural decisions related to data, modeling, and deployment in collaboration with specialized teams Strong leadership and communication skills to influence engineering culture without direct authority.
Deep understanding of aviation systems including PNRs, e-tickets, and revenue management logic (Yield and Inventory control).
Proficient in MS Office.
Work Experience With a minimum of 6-8 years of experience in Data Sciernce, MLOps, Platform Engineering, or DevOps specifically for machine learning.
Out of which a minimum of 1-2 years of experience in the aviation domain • airline, vendor.
Strong experience designing scalable and impactful solutions Hands-on experience with Snowflake/Snowpark and Airflow.