Piramal - Assistant Vice President - Model Monitoring & Validation - Risk Analytics

Piramal - Assistant Vice President - Model Monitoring & Validation - Risk Analytics (6-11 yrs)

Role & responsibilities :

1. Model Performance Monitoring :

- Own and operate the end-to-end model monitoring framework for all credit, collections, fraud, and pricing models in production.

- Track key model health metrics including PSI, CSI, Gini coefficient, KS statistic, AUC drift, rank-order stability, and vintage performance on a regular cadence.

- Design and implement automated monitoring pipelines that trigger alerts when model performance degrades below defined thresholds.

- Build scorecard monitoring dashboards integrating bureau, internal, and alternate data to provide a real-time view of model health.

- Conduct challenger vs. champion analysis to recommend model upgrades, recalibrations, or full redevelopments.

2. Model Risk Governance & Inventory Management

- Maintain a comprehensive model inventory covering model purpose, ownership, tier classification, deployment date, last validation date, and refresh schedule.

- Define and enforce model risk tiering (High / Medium / Low) based on materiality, complexity, and regulatory impact.

- Establish and oversee the model lifecycle governance process from development approval through deployment, monitoring, and decommissioning.

- Lead periodic model validation reviews in coordination with model owners and independent validators; ensure findings are tracked to closure.

- Drive adherence to RBI model risk guidelines and internal model risk policy, ensuring audit-readiness at all times.

3. Reporting, Insights & Stakeholder Leadership

- Produce regular model health reports for Senior Leadership, the Risk Committee, and the Board translating technical metrics into clear business narratives, and leveraging generative AI to automate monitoring commentary and board-ready model risk updates.

- Build executive-level model risk dashboards surfacing portfolio-wide model performance, open validation findings, and upcoming refresh obligations across all risk tiers.

- Liaise with CRO and Regulators to respond to model-related queries, maintain a model exception register, and ensure all threshold breaches are documented, escalated, and resolved.

- Mentor and develop a team of model monitoring analysts and data scientists in statistical testing, MLOps platforms and responsible AI governance best practices.

- Partner with Model Development, Credit Risk, Technology, and Data Engineering teams to embed monitoring at every stage of the model lifecycle and champion a culture of model risk and responsible AI across the organisation.

Ideal Candidate Profile :

Education: Bachelors/Masters degree in a quantitative or technical discipline (Economics, Statistics, DataScience, Computer Science, Finance, Engineering)

Experience : 6-8 years of experience in analytics, ideally in credit risk analytics/financial services with exposureto secured portfolios and AI-driven decision systems.

- Proficient in Python, R, SQL, SAS, and hands-on experience with ML libraries (scikit-learn, TensorFlow,PyTorch, XGBoost).

- Deep understanding of credit scoring, bureau data, and model validation techniques (PSI, CSI, Gini, KS)

- Experience with vintage analysis, roll-rate analysis, and loss forecasting

- Understanding on AI fairness, bias auditing, and explainability testing including disparate impact analysisand counterfactual evaluation

- Understanding of Responsible AI principles and emerging AI governance regulations (RBI AI guidelines)

Why this Role matters :

- Every model-driven decision from loan origination to collections strategy rests on the assumption that the modelis working as intended. As AI becomes the core engine of our credit intelligence, this role is the guardian of thatassumption at scale.

- Ensure AI and ML-driven credit decisions remain accurate, fair, and defensible as portfolios evolve andmacroeconomic conditions shift.

- Build the AI monitoring infrastructure that allows the business to deploy models at scale without compromising riskquality or regulatory trust.

- Create regulatory confidence by maintaining a transparent, auditable, and fully documented model governanceecosystem meeting the bar for RBI, and future AI regulation.

- Protect the organisation from model failure-induced losses through AI-powered early detection, automated alerts,and proactive model refresh.

- Champion Responsible AI ensuring every model we deploy is explainable, fair, and free from unintended bias.

- Enable the next generation of AI innovation by establishing the governance rails that make experimentation safe,scalable, and trusted.