Sr. Applied Scientist - AI Velocity Team, Applied AI Acceleration Solutions Architecture

About the role

Are you passionate about using data science to transform how businesses understand and optimize customer interactions at scale? Do you want to build the models and analytics that power the next generation of AI-driven customer experiences while working directly with customers to accelerate production deployments? As a Senior Applied Scientist within the Applied AI Solutions team, you will collaborate across AI Velocity Teams (AIVT), enabling multiple customer engagements simultaneously. You will lead data science initiatives that span the full lifecycle - from identifying high-value business problems and formulating hypotheses, through rigorous experimentation and modeling, to deploying production-grade solutions that serve thousands of customers. You will bring deep expertise in statistical inference, machine learning, and experimental design to drive measurable impact across Amazon Connect's analytics products and broader Connect AI initiatives. A critical dimension of this role is working directly with customers during production pilots to accelerate time-to-value. You will partner with Applied AI Solutions Architects and Customer Success Specialists to design, build, and deploy AI solutions in customer environments during fixed deployment cycles. You will enable field teams with data-driven insights, reusable analytical assets, ROI tools, and scalable tooling that accelerate customer engagements and solution delivery. Your work will directly influence customer decisions to adopt Connect Customer AI by quantifying business outcomes and demonstrating measurable value. You will operate with significant autonomy, owning the scientific direction of your projects while collaborating with applied scientists, software engineers, product managers, technical, and business stakeholders. You will be expected to identify the right methodology for each problem - whether that's a classical statistical approach, a modern deep learning technique, or a novel combination - and communicate your findings clearly to both technical and non-technical audiences. This role spans Connect AI initiatives including conversational analytics and agentic AI capabilities, offering the opportunity to pioneer data science approaches that scale intelligent analytics worldwide. Key job responsibilities - Design, develop, and deploy statistical models and machine learning pipelines to drive product improvements and business decisions - Work directly with customers during production pilots to design, build, and deploy AI solutions that demonstrate measurable business value - Design and execute A/B experiments and causal inference analyses to measure the impact of new features and model changes on customer outcomes - Build ROI models and business case tools that quantify the value of Connect Customer AI for existing customers transitioning from Connect Customer Basic - Develop and maintain forecasting systems for demand prediction, capacity planning, and workforce optimization - Develop and apply NLP and generative AI techniques to extract insights from structured and unstructured data at scale - Partner with applied scientists and software engineers to productionize models, ensuring reliability, monitoring, and operational excellence - Enable AI Velocity teams with reusable analytical assets, diagnostic notebooks, and scalable tooling that accelerate customer engagements - Build benchmarking studies and optimization frameworks that demonstrate value across customer cohorts - Own success metrics and create mechanisms to measure model performance, adoption, and business impact - Communicate findings and technical trade-offs to senior leadership and customer executives through written documents (6-pagers, science reviews) and presentations - Operate as a shared resource across 2-3 AIVT teams simultaneously, providing data science expertise across multiple customer engagements A day in the life - Start the morning on a call with the AI Velocity Teams preparing for a strategic customer engagement - reviewing the analytical assets and dashboards you've built, walking through how to interpret model outputs, and tailoring recommendations to the customer's contact center environment - Join a customer working session where you're deploying a production pilot - analyzing their historical contact data, building demand forecasting models, and demonstrating how AI optimizations will reduce their cost per serviced contact while improving customer experience metrics - Dive into a deep analysis triggered by AIVT field feedback - a large enterprise customer is seeing unexpected patterns in their contact data, and you're pulling together multi-source data to isolate root cause and build a reusable diagnostic notebook the AIVT team can leverage for similar cases - Participate in a Conversational Analtyics science review, presenting your A/B test results on a new sentiment classification approach and discussing trade-offs between model acc