MLOps Architect: Lead Game AI Platform

EPAM Systems; Colombia; Remote; COP 287,954,000 - 415,933,000; Full time

Benefits offered by this job

Healthcare benefits

Paid time off

Learning opportunities

Job summary

EPAM Systems, Inc. in Colombia seeks a Lead MLOps Engineer to head an MVP engagement for a major AAA game publisher. You will design a reproducible platform with full lineage from day one and ensure production readiness across ML components.

You will mentor engineers, define standards for model lifecycle, provenance, and calibration harness, and drive cross‑franchise collaboration to deliver scalable, auditable ML systems.

Qualifications

• 5+ years in MLOps or ML platform engineering.
• Deep expertise in ML model lifecycle management, including versioning and rollback.
• Strong background in architecting signal computation pipelines.
• Advanced proficiency in calibration methodology and holdout discipline.
• Advanced skills in Python and SQL for ML pipeline automation.

Responsibilities

• Define the MLOps architecture and strategy for the platform across franchises.
• Own signal catalogue operations end-to-end and define provenance standards.
• Set strategy for Semantic Vector Index versioning and controlled reindex paths.
• Design and own the Back-test & Calibration Harness with as-of temporal filtering.
• Lead model generation, experimentation framework, and cross-franchise governance.
• Mentor engineers and review designs, representing MLOps in cross-team discussions.

Skills

MLOps leadership

Python

SQL

Model monitoring

Drift detection

Experimentation framework

Cross-functional collaboration

Mentorship

Versioning

Back-test harness

Tools

Snowflake Snowpark

pgvector

Job description

EPAM Systems, Inc. in Colombia seeks a Lead MLOps Engineer to head an MVP engagement for a major AAA game publisher. You will design a reproducible platform with full lineage from day one and ensure production readiness across ML components.

You will mentor engineers, define standards for model lifecycle, provenance, and calibration harness, and drive cross‑franchise collaboration to deliver scalable, auditable ML systems.