• Architected and deployed an AI-driven incident intelligence platform for ServiceNow ITSM using Python, PyTorch, and PySpark to automatically classify, prioritize, and route enterprise support tickets, reducing manual triage workload by 45% across high-volume IT operations.

• Built large-scale ML data pipelines using Apache Kafka, Airflow, and Spark to ingest and process 12M+ historical incidents and 500K monthly service requests, enabling real-time feature generation for predictive incident management models.

• Developed Transformer-based NLP models (BERT/Finetuned LLMs) to analyze ticket descriptions and recommend knowledge base resolutions, improving automated incident classification accuracy by 28% and reducing average resolution time by 32%.

• Implemented predictive outage detection models using time-series ML techniques and gradient boosting to analyze infrastructure metrics and historical incidents, enabling early detection of critical system failures and reducing major incident occurrences by 20%.

• Designed a vector-based knowledge retrieval system using Sentence Transformers and OpenSearch to match incoming incidents with similar historical cases and remediation steps, increasing first-response resolution effectiveness by 30%.

• Integrated RAG-based generative AI assistants using Azure OpenAI APIs and enterprise knowledge bases to automatically summarize incident tickets and generate troubleshooting recommendations, accelerating support engineer response time by 35%.

• Productionized ML models through containerized deployment on Kubernetes (EKS) with MLflow-based experiment tracking and CI/CD pipelines, ensuring scalable inference services supporting millions of monthly ITSM transactions with >99.9% availability.

• Developed and deployed predictive models on large scale HR data to forecast attrition, retention, optimize recruitment and improve workforce planning through scalable ML solutions and interactive dashboards.

• Built predictive models (logistic regression, random forests, XGBoost) to forecast attrition, identify high performers and estimate recruiting demand, directly supporting strategic workforce planning.

• Led end-to-end analytics and ML initiatives to optimize the hiring pipeline, reducing time-to-fill and improving candidate experience through datadriven insights and measurable business impact.

• Queried, cleaned, and transformed large-scale HR datasets using SQL and Apache Spark, accelerating feature engineering, model development, and analytical workflows.

• Designed and executed A/B tests and statistical experiments to evaluate hiring strategies, improving recruiter effectiveness and conversion rates.

• Deployed machine learning models using FastAPI, Docker, and MLflow on AWS (EC2, Lambda, S3), while collaborating with BI and data engineering teams to strengthen data pipelines, automation, and data quality, supported by Git-based version control.

• Developed interactive dashboards using Tableau and Power BI to track key HR metrics like turnover, hiring efficiency and productivity which helped non-technical stakeholders in making data-driven decisions.

• Trained junior analysts and guided them through data preprocessing, choosing the right models, optimization and building dashboards.