Sarp Coskun
Leading builders at scale to ship AI-native advertising products, shaping what comes next in ad tech.
About Me
I build advertising products. For more than a decade at Amazon Ads, I have led the people who turn targeting, measurement, and in-store experiences into products advertisers actually run. Multi-agent systems and GenAI are part of how we ship, not a side project.
As Sr. SDM, I lead product builders behind intelligent ad targeting in Display Ads, including the targeting agent we launched at unBoxed. Before that I ran a 35+ person organization across Predictive Audiences, Events Manager, and the measurement pipes the business stands on: Conversions API, Amazon Ad Tag, and MMPs. Earlier products include in-store attribution for Whole Foods Market and Amazon Fresh, Sponsored Display in nine countries, and patented work in physical-world advertising.
I'm also an Adjunct Professor of Computer Science at Bellevue College, where I teach Deep Learning and Data Science and serve on the CS Industry Advisory Board. The through-line is the same in both rooms: pick a hard product problem in ad tech, put a strong team on it, and ship something the market can feel.
My Work
Engineering Leadership
Amazon / Amazon Ads
- Sr. SDM & Head of Engineering: Led multidisciplinary organizations of 35+ EMs, TPMs, scientists, and full-stack engineers to build core intelligent ad targeting experiences and large-scale measurement signal ingestion pipelines (Conversions API, Amazon Ad Tag, MMPs).
- Physical Realm Advertising: Spearheaded full-stack engineering and ML systems powering in-store ad attribution and digital signage for Whole Foods Market and Amazon Fresh; filed patent for innovative ad technology.
- Sponsored Display & Platform Scalability: Scaled multi-channel display advertising globally across 9 countries, driving high-throughput microservices, sub-second latency APIs, and robust cross-team developer tooling.
- GenAI & MLOps Practices: Pioneered GenAI-native leadership practices, automated developer workflows, and scalable predictive ML pipelines for automated audience modeling.
Research Assistant
Case Western Reserve University (Data Research Lab)
- Interactive Research Tools: Designed and implemented online web applications for systems biology researchers to simulate, compose, and compare complex mathematical models via SOAP web services.
- Data Infrastructure & DevOps: Maintained research platforms and data lakes across multi-server environments; established Git version control and automated testing pipelines (NUnit, Selenium) to increase deployment reliability.
- Academic Mentorship: Supervised graduate students on software engineering best practices, automated testing, and database architecture.
Software Engineer
L-Mobile Solutions
- Enterprise CRM Platform: Engineered a new web-based, multi-language CRM application for healthcare sector clients within a distributed engineering team.
- Architecture & Quality: Applied modern MVC design patterns and built extensive automated unit testing suites to maximize testability and long-term code maintainability.
Courses Taught
Bachelor of Science in Computer Science, Data Science Emphasis at Bellevue College
CS 465 - Deep Learning
An introductory exploration of deep learning technologies in AI. Covers theoretical foundations of neural networks and practical experience with leading open-source tools like Keras and PyTorch.
CS 310 - Python for Data Science
Hands-on experience solving real data science challenges, including modeling, statistics, and storytelling using popular libraries like Pandas, NumPy, Matplotlib, and Scikit-Learn.