George Petre

• Shipped Generative AI experiences for Amazon Customer Reviews and helping customers make unregrettable purchase decisions on the Amazon Marketplace (highlighted by mainstream media e.g., CNN, CNBC, BusinessInsight). • Owned processes to keep senior leadership engaged and align multi-functional teams (Legal, Tech, MLOps, CX, PR) with navigating at high speed in the AI ambiguously regulated space. • Optimized supply chain metrics (e.g., out-of-stock, availability) by adjusting complex interconnected systems

Leading adidas app expansion in over 40 countries, growing revenue from 0 to 10 digits and an active user base from 0 to 8 digits. Team & Partners accountabilities: • Managed & coach people across all mobile & digital team contractors and enterprise partners • Managed internal stakeholders across the globe ( US, EU, CN, RU, PAC)

As a Product Manager I was in charge of creating products that make private users, small and medium businesses successfully in online environment. Main Responsibilities: - setup the mobile knowledge inside the 1&1 Hosting division by creating the first strong team specialized in Mobile Applications for iOS and Android. - defined specifications for a facelift of all the current existing apps, previously developed by third parties, by improving Flows and Visual Design

Creating innovative security products for the dynamic social media environment

As a technical team leader, I was in charged with the recruitment, coaching and management of a small technical team focused on developing honeypots. Honeypots are systems used to collect samples of malware/spam/phishing/other threats in order to train automatic filters and to understand better the vulnerabilities.

As part of the Online Threats Team, I was in charged with the honeypot development and threat research. Since honeypot development require complex networking configuration, I was also in charged with the management of Online Threats IT infrastructure.

As one of the first members of the antispam team, I was in charged with the development of e-mail honeypots in order to create a relevant (world scale) spam. Shortly, the system started to receive millions of e-mails. Since the filters should be trained as fast as possible, I have also developed an scalable infrastructure, able to send in real time this samples to different training processes.

Mathematics and Computer Science