• Lead the development + implementation of real-time feature detection and anomaly detection models • Generate data characteristic requirements for real-time data processing pipelines • Prepare technical documentation, reports, and specifications • Collaborate with cross-functional teams including project managers, technicians, and other engineers • Perform testing, troubleshooting, and quality assurance on systems or products • Ensure compliance with safety regulations, industry standards, and company policies

• Build, deploy, and optimize the machine learning services that power the Tebra platform • Write high-quality, production-grade software for data ingestion, feature extraction, and model inference • Implement robust CI/CD pipelines, automated testing, and comprehensive logging for deployed models • Construct and maintain specific data pipelines for training and inference • Develop reusable software modules and utilities to streamline the development process • Translate business requirements into technical specifications • Monitor the daily performance of production models and debug incidents

• Translate complex business problems into solutions that leverage end-to-end, production-level machine learning systems • Develop, iterate on, and productionize machine learning models, owning both model logic and supporting systems. • Implement and own backend pricing services and APIs consumed by the Lime App. • Architect and build scalable, reliable platforms and services that power pricing decisions and enable new product initiatives. • Build trust with leadership to influence the company's pricing strategy through reasoning and communication of model performance and technical vision • Collaborate with other Software Engineers and Data Scientists to raise the bar on ML standards and Engineering Excellence • Mentor and support engineers on the team, helping grow technical depth and develop future leaders.

• Architect the Hybrid Engine: Design and build the "Project Bridge" architecture, transitioning the platform from heuristic-based logic (Cohort/Geo-based) to fully real-time ML personalization (Vector Search/Neural Networks). • Real-Time Inference at Scale: Steer the design and deployment of low-latency services (Segment Service & User Profile Service) using Redis/DynamoDB to serve personalized board orderings, deposit defaults, and "For You" feeds in milliseconds. • Feature Engineering & Data Strategy: Partner with Data Science to build the logging pipelines that tag why a user saw an item (data labeling). You will create the feature store required to train future neural networks for individual-level personalization. • Solve the "Cold Start" Problem: Implement logic for dynamic league ordering and deposit smart-defaults based on geospatial data and initial user cohorts, ensuring immediate relevance for new users.