Interview : Video
Visa : USC, GC
This is hybrid from day-1
Description :
This is a greenfield role and needs someone that has come from environments doing the same thing, building from scratch
Our last candidate for this role, Ekshitha, did not have practical hands-on experience with Snowflake MLOps and CI/CD. This experience is very important. She also did not demonstrate being able to work as an individual. Seemed like she mostly worked in a team setting. This role will require being the sole GoTo person as they start off.
POSITION SUMMARY:
Republic Services is building its first enterprise-scale MLOps platform on Snowflake. This role is a rare opportunity to design,
build, operate, and continuously mature the end-to-end ML production ecosystem from scratch. The platform will be built on
top of the enterprise medallion architecture (Bronze, Silver, Gold), and ML models will consume curated, governed data
products from these layers to ensure quality, lineage, and scalability.
RESPONSIBILITIES:
Architect and build a production-grade MLOps platform on Snowflake using Snowpark, Snowflake ML, Model Registry,
and Feature Store capabilities.
Design and operationalize reusable ML pipelines for training, validation, deployment, inference, and monitoring.
Build MLOps workflows aligned to Bronze, Silver, and Gold layers so model training and inference consistently
consume trusted medallion data.
Establish model lifecycle management standards, including versioning, approval workflows, promotion gates, rollback
strategy, and model lineage.
Partner with data scientists to productionize models quickly and safely, transforming experiments into reliable,
scalable services.
Implement model observability for performance, drift, bias, data quality, and service reliability with actionable
alerting and SLOs.
Automate retraining and refresh workflows using Snowflake Tasks, Dynamic Tables, and event-driven orchestration
patterns.
Partner with data engineering to ensure feature pipelines are reliable, reusable, and synchronized with medallion
layer evolution.
Define and implement CI/CD for ML workflows (code, data, models, and configuration), including testing frameworks
and release controls.
Drive MLOps governance across security, compliance, auditability, reproducibility, and responsible AI practices.
Lead platform maturation from MVP to enterprise scale, including documentation, developer enablement, and
operational runbooks.
REQUIRED QUALIFICATIONS:
5+ years of experience in ML Engineering, MLOps, or related platform engineering roles.
Strong Python and SQL expertise, with proven experience building production ML pipelines.
Hands-on experience with Snowflake data and compute patterns; experience with Snowpark and Snowflake-native ML
tooling strongly preferred.
Demonstrated experience with model deployment, versioning, monitoring, and lifecycle governance in production.
Experience implementing CI/CD and testing strategies for ML systems.
Solid understanding of feature engineering pipelines, training-serving consistency, and data quality controls.
Experience with cloud infrastructure and services (AWS preferred).
Strong collaboration skills and ability to work cross-functionally with data science, data engineering, and business
stakeholders.
PREFERRED QUALIFICATIONS:
Experience with Snowflake Model Registry, Snowflake Feature Store, and model observability within Snowflake.
Experience designing ML systems on medallion/lakehouse-style data architectures.
Experience with dbt or similar transformation frameworks.
Familiarity with streaming or near-real-time inference patterns.
Experience in high-volume operational domains such as logistics, fleet, route optimization, or environmental services.
Prior experience building greenfield platforms and defining operating standards from the ground up.