ML Operations Engineer – Lead

Positions: 2 (full-time)

Experience: 18–25 years

Job title: ML Operations Lead

Job level: E1 (Senior Manager)

Division: AI & Agent Operations (Day 2), AI Organisation

Reports to: Assistant Director, AI & Agent Operations (Singapore)

Location: Hinjewadi Phase 1, Pune, India

Work mode: 5 days a week in the office

Work hours: 9:00 a.m.–6:00 p.m. IST

Direct reports: 2 ML Operations Engineers (India)

Role focus: Traditional data science and machine-learning models, not LLMs or generative AI.

Position Summary

Lead the machine-learning operations team in India, operating the organisation’s traditional data science and ML models in production. Examples include churn prediction, fraud and revenue assurance, and network capacity forecasting across consumer, enterprise (B2B), and network businesses. Own the practices that keep predictive models accurate and safe in production, including drift monitoring, retraining and validation, deployment and rollback, lineage, and cost management. Lead a team of two engineers and coordinate with the AI & Agent Operations lead in Singapore on changes, incidents, and standards. This is a classical MLOps role.

Key Responsibilities

1. Lead and develop the India ML operations team; set direction, standards, and priorities for the model estate. Translate business requirements and communicate them clearly to stakeholders. Experience with churn models is preferred.
2. Own deployment and retraining practices, including backtesting or champion-challenger validation gates before promoting models, and a rehearsed rollback path.
3. Own drift and performance monitoring across the portfolio, including data, feature, and concept drift, as well as model-specific retraining triggers and thresholds.
4. Maintain the lineage registry, including version, training data, baseline, and last-known-good model, and ensure model auditability.
5. Deliver business-requested model changes within agreed timelines. Monitor bias and fairness in decision-making models, such as churn and fraud models, and keep humans in the loop where appropriate.
6. Manage the partnership with Data Engineering, including data contracts, freshness, and change notifications; treat data-quality issues as a priority.
7. Coordinate with the Singapore lead and team on change scheduling, incidents affecting shared services, and estate-wide standards.
8. Report model health, deployment success, and portfolio performance against the function’s KPIs.

Challenges

• Running the model estate from India while the overall function is led from Singapore, requiring cross-border coordination on standards, changes, and incidents.
• Managing drift, model decay, retraining discipline, and data-quality dependencies across a portfolio with a lean team of three.
• Balancing model performance with fairness and transparency for models that inform customer and revenue decisions.

Decision-Making Authority

Decisions made within the role:

• Retraining and model-promotion decisions within the validation-gate discipline.
• India team priorities, standards, and drift-response actions.
• Rollback to a previous model or safe baseline for models operated by the team.
• Data-quality escalations to Data Engineering.

Decisions referred to higher authority:

• Cross-functional priorities and estate-wide standards (Singapore lead).
• Enterprise risk acceptance and regulatory considerations concerning model use.
• Headcount, budget, and procurement beyond delegated limits.
• Changes affecting wider Day 1, data, or platform functions.

Qualifications, Experience, and Skills

Education and qualifications

• Degree in Computer Science, Data Science, Engineering, Statistics, or a related field.
• Postgraduate qualification in ML/AI or a quantitative discipline is preferred.

Work experience

• 8+ years in ML, MLOps, or data-science engineering, including team leadership.
• Hands-on experience deploying and operating traditional ML models in production.
• Experience with drift monitoring, retraining, and model validation.
• Telco-scale data or model environments, customer churn, team management, cross-border coordination, and business stakeholder engagement are preferred.

Technical and professional skills

• MLOps, including CI/CD for models, model registries, and lineage.
• Drift detection (data, feature, and concept) and retraining pipelines.
• Champion-challenger testing, backtesting, and validation practices.
• Model performance metrics, including AUC and precision-recall, and model decay management.
• Bias and fairness monitoring.
• Cloud ML platforms.

Leadership and interpersonal skills

• People leadership and mentoring.
• Disciplined operational judgment.
• Clear communication across a cross-border reporting line.
• Stakeholder management with data and business teams.

Other relevant knowledge

• Understanding of telco predictive use cases, such as churn, fraud, revenue assurance, and network forecasting.
• Data governance and model-risk considerations.

Notice period: Immediate joiners and candidates currently serving notice are welcome.