Lead AI-ML Engineer

Key Responsibilities:

• Collaborate with stakeholders to understand business objectives and define requirements for anomaly detection.
• Develop, optimize, and maintain computational models for debit transaction anomaly detection using AI/ML techniques.
• Perform data analysis, generate insights, and identify patterns to support decision-making.
• Design and implement statistical models, including standard deviation calculations, variance thresholds, and probabilistic models to enhance anomaly detection accuracy.
• Work with existing models to apply backtracking methodologies and improve anomaly reduction strategies.
• Leverage machine learning algorithms (e.g., classification, clustering, time-series modeling) to predict, detect, and manage anomalies.
• Collaborate with engineers and business teams to integrate models into production systems.
• Conduct performance monitoring, fine-tuning, and validation of ML models to ensure accuracy and reliability.
• Prepare technical documentation, visualizations, and reports to communicate findings effectively to business and technology stakeholders.

Required Skills & Qualifications:

• Bachelor’s or Master’s degree in Computer Science, Data Science, Statistics, Mathematics, or a related field.
• 10+ years of hands-on experience in data science, AI, or ML engineering.
• Strong proficiency in Python, R, or Scala with experience using data science libraries (e.g., NumPy, Pandas, Scikit-learn, PyTorch, TensorFlow).
• Solid understanding of Data Science with a heavy focus on statistical modeling and Machine Learning, hypothesis testing, regression analysis, and variance modeling.
• Experience with anomaly detection techniques — supervised, unsupervised, and hybrid approaches.
• Experience in Generative AI based implementations.
• Expertise in working with large datasets using SQL, Spark, or similar data-processing frameworks.
• Strong problem-solving, analytical thinking, and communication skills.
• Experience in deploying ML models into production environments, MLOps, preferably on AWS