AdTech
summary
Mobile ad platform handles 1.5 billion devices and processes 8–10 billion ad requests/day, using deep neural nets to operate in microseconds. Concrete wins include anomaly detection that stops fraud before scale, models that output creative feature vectors to predict click rates, and persona-style real-time segmentation from high‑dimensional unstructured signals (images, audio, sensors). Teams report hundreds of use cases internally and a working zero-shot creative scoring pipeline that predicts ad performance without prior exposure.
AI-written summary · about
chapters (6)
- 0:00Start
- 14:34The Fraud Algorithm
- 20:01What's Possible with in the World of Ai Meets Creativity
- 23:10Image Recognition
- 26:57The Zero Moment of Truth
- 31:20When Does Advertising Become Intrusive and How Do You Respect User Preference
related talks
29mAI and ML for Computational Advertising
37mSupercharging CTR (Click-Through Rate) with Plug-and-Play AI Capabilities
24mKeynote: AI Trends in Performance Marketing
04mCES Interview
33mMachine Learning to moderate ads
23mThe unreasonable effectiveness of data for better monetisation
32mAI for Mobile and IoT Devices: TensorFlow Lite
37mApplying ML in AdTech and Lifecycle of an ML project