Applied Data Scientist

Role Overview: You will be part of a team at Microsoft that focuses on building the intelligence layer for the next-generation threat detection ecosystem. This involves using deep applied science, graph-theoretic reasoning, large-scale machine learning, and multi-modal security analytics to uncover hidden attack patterns across identity, endpoint, network, and cloud. Your work will drive advancements in attack-path discovery, anomaly detection, graph construction, and threat-hunting experiences within Microsoft Security. Key Responsibilities: - Develop supervised and unsupervised machine learning models for anomaly detection, fraud/threat pattern discovery, alert classification, confidence scoring, and signal fidelity improvements. - Build and maintain feature pipelines over multi-modal security telemetry including identity, endpoint, network, and cloud. - Apply graph-focused machine learning techniques such as graph embeddings, GNNs, similarity scoring, and relationship modeling. - Contribute to graph construction logic, schema evolution, and ontology-driven enrichment for various components like Verdict Net, Verdict Propagation, Campaign Graphs, and Vortex insights. - Implement graph traversal, multi-hop reasoning, and cluster detection algorithms to surface hidden attack patterns. - Analyze large, noisy, high-dimensional security datasets using tools like ADX/Kusto, Spark, and distributed compute platforms. - Run A/B experiments, offline evaluations, and benchmark models to continually enhance detection quality. - Collaborate with detection engineering, threat research, product teams, and red teams to integrate machine learning outcomes into real-world protection experiences. - Translate complex analytical insights into actionable improvements for detections, disruptions, and customer-facing intelligence. Qualifications: - Bachelors degree in Computer Science, Data Science, Electrical Engineering, Mathematics, or a related field AND a minimum of 6 years of hands-on experience in data science and machine learning. - Strong proficiency in Python, machine learning frameworks such as PyTorch/TensorFlow, and data processing libraries. - Experience with machine learning techniques including gradient-boosted models, supervised/unsupervised learning, embeddings, clustering, and anomaly detection. - Experience querying and analyzing large datasets using tools like Kusto, SQL, Spark, or equivalent data engines. - Strong fundamentals in probability, statistics, and algorithmic thinking. - Ability to write clean, reliable research code and effectively communicate findings. Note: This position will be open for a minimum of 5 days, with applications accepted on an ongoing basis until the position is filled. Microsoft is an equal opportunity employer. Role Overview: You will be part of a team at Microsoft that focuses on building the intelligence layer for the next-generation threat detection ecosystem. This involves using deep applied science, graph-theoretic reasoning, large-scale machine learning, and multi-modal security analytics to uncover hidden attack patterns across identity, endpoint, network, and cloud. Your work will drive advancements in attack-path discovery, anomaly detection, graph construction, and threat-hunting experiences within Microsoft Security. Key Responsibilities: - Develop supervised and unsupervised machine learning models for anomaly detection, fraud/threat pattern discovery, alert classification, confidence scoring, and signal fidelity improvements. - Build and maintain feature pipelines over multi-modal security telemetry including identity, endpoint, network, and cloud. - Apply graph-focused machine learning techniques such as graph embeddings, GNNs, similarity scoring, and relationship modeling. - Contribute to graph construction logic, schema evolution, and ontology-driven enrichment for various components like Verdict Net, Verdict Propagation, Campaign Graphs, and Vortex insights. - Implement graph traversal, multi-hop reasoning, and cluster detection algorithms to surface hidden attack patterns. - Analyze large, noisy, high-dimensional security datasets using tools like ADX/Kusto, Spark, and distributed compute platforms. - Run A/B experiments, offline evaluations, and benchmark models to continually enhance detection quality. - Collaborate with detection engineering, threat research, product teams, and red teams to integrate machine learning outcomes into real-world protection experiences. - Translate complex analytical insights into actionable improvements for detections, disruptions, and customer-facing intelligence. Qualifications: - Bachelors degree in Computer Science, Data Science, Electrical Engineering, Mathematics, or a related field AND a minimum of 6 years of hands-on experience in data science and machine learning. - Strong proficiency in Python, machine learning frameworks such as PyTorch/TensorFlow, and data processing libraries. - Experience with machine learni