Staff Machine Learning Engineer, Abnormal Security

• 8+ years of experience designing and building high-impact, customer-facing machine learning applications
• Proven experience working on ML at scale with direct product impact in mature ML industries such as recommendation systems, ad tech, quantitative finance, or fraud detection
• Strong grasp of the theoretical limitations of deep learning models and a systematic approach to investigating and debugging poor model performance
• Demonstrated experience in the productionization of large-scale ML models in fast-feedback environments
• Ability to reason about abstract system gaps and propose generalizable, architecturally sound ML solutions, not just point fixes
• Expertise across the entire ML lifecycle, from data exploration and feature engineering to model deployment and online scoring
• Fluency in Python and ML frameworks like Scikit-learn, PyTorch, or TensorFlow
• BS degree in Computer Science, Applied Sciences, Information Systems, or a related engineering field
• This position involves access to technology that is subject to the U.S. Export Administration Regulations (EAR). As a result, candidates offered employment must be eligible to access controlled technology under U.S. export control laws. Employment in this position is conditioned on the availability of government authorization

Desirable

• Abnormal Security is looking for a Staff Machine Learning Engineer to join the Message Detection - Attack Detection team
• In a landscape where a single successful attack can lead to financial losses of millions of dollars, the Attack Detection team plays the central role of building an extremely high recall Detection Engine that can operate on hundreds of millions of messages at milliseconds latency
• The Attack Detection team’s mission statement is to provide world-class detector efficacy to tackle changing attack landscape using a combination of generalizable and auto trained models as well as specific detectors for high value attack categories
• This role is central to our mission of protecting the world’s largest enterprises
• You will be responsible for reasoning about the gaps in our multi-layered detection system and proposing generalizable ML solutions
• You will have a significant impact on our technical roadmap, guiding how our diverse set of detection models—spanning behavioral analysis, natural language understanding, and deep learning systems—work in concert
• This is a unique opportunity to shape the future of our ML architecture, from evolving our core training and deployment strategies for a global infrastructure to defining how our core ML capabilities can be exposed as scalable services to power other products across the company
• Serve as a technical leader and subject matter expert, providing architectural guidance and mentorship across multiple machine learning workstreams
• Architect and design generalizable ML systems to address the most critical gaps in our detection capabilities, moving beyond incremental improvements
• Reason holistically about our entire detection engine, defining the architectural vision for how different classes of models—from heuristic and behavioral to complex deep learning systems—should integrate and operate
• Drive the technical roadmap for foundational, long-term projects, such as evolving our global model training paradigms and creating centralized ML capabilities that can be leveraged as platforms by other teams
• Provide technical mentorship and feedback on ML decisions across different workstreams, elevating the performance of the entire team
• Own the end-to-end ML lifecycle: from data analysis, feature engineering, and model prototyping to working with infrastructure teams on productionization, deployment, and monitoring of large-scale models
• Investigate complex model performance issues, applying a deep theoretical understanding of machine learning and deep learning to diagnose and resolve them
• Continuously adapt our systems to new, unseen attacks by developing and refining our automated model retraining and evaluation pipelines