The longer version. For the short version, see About.
1996. Started programming at age 12. Visual Basic, early web development. The beginning of a lifelong habit of taking things apart to see how they work.
2009. Graduated from Bocconi University in Milan with an MSc in International Management (110/110, the maximum grade). CEMS MIM exchange at Aalto University in Helsinki. Graduate thesis: “Commercializing Radical Technologies.”
2010. Founded Discontinuity S.r.l., a software consultancy based in Northern Italy. Over the next eight years, the company worked with clients ranging from Honeywell (IoT projects) to Gucci (mobile experience), and developed machine learning tools for Italian insurance companies. The consultancy remained profitable for the entire period.
2011. Enrolled in Andrew Ng’s Stanford Machine Learning class (August 17, 2011). This was before Coursera existed as a platform and a year before AlexNet triggered the deep learning revolution. The class was one of three Stanford courses that launched the modern era of online education.
2012. Co-authored CocoaPods, which became the standard dependency manager for iOS development. Wrote the dependency resolver, the plugin system, and the CDN infrastructure. The project grew organically without marketing because it solved a real problem for iOS developers. It eventually outgrew its original authors and is now maintained by an active community of contributors.
CocoaPods today: over 3 million apps, 100,000+ libraries, over 100 million RubyGems downloads. Used by Amazon, Google, Facebook, Uber, and Slack. Referenced in React Native’s iOS setup documentation. Google built developer content around it.
2014. Keynoted at a WWDC community event on CocoaPods and the iOS dependency management ecosystem.
2015. Listed machine learning, artificial intelligence, and deep learning as core professional interests. Began following Andrej Karpathy’s work on neural networks and his “Software 2.0” thesis closely.
2016. A pivotal year for connecting the dots between AI and business:
Published a blog series on technology foresight, including:
- “Management by exception”: described automated workflows that only escalate to humans when something breaks the pattern. This is essentially the architecture that Intarsia uses today.
- An analysis of the economy of algorithms and how ML would reshape business operations.
- A piece predicting Apple’s eventual switch from Intel to ARM chips. Apple announced the transition four years later, in 2020.
Described Discontinuity as “a startup whose core offerings are AI technologies” in a company blog post. This was years before applied AI became mainstream.
2016 (cont). Built the Mowis platform with Sintex, a Siemens automation partner: a supervisory control system (SCADA) that monitors and controls plastics manufacturing lines in real time across factories worldwide for Moretto. Replaced manual gauges and paper logs with a unified digital dashboard. The system is still in production today.
2017. Selected as an Intel Software Innovator for AI/ML. Through the program, presented “The Economic Implications of Machine Learning” at the Intel Nervana AI Academy meetup in Rome (November 30, 2017). The talk argued that:
- ML automation would evolve through four stages: basic automation, human amplification, autonomous execution with human supervision, and full autonomous execution.
- The true power of the technology was enabling the wider population, not replacing specialists.
- Software written by neural networks (what Karpathy called “Software 2.0”) would eventually be easier to produce than hand-written code.
All three predictions have since been validated by the emergence of agentic AI platforms, human-in-the-loop automation, and tools like GitHub Copilot and Claude Code.
2018. Joined Amazon. First year at AWS, working on internal mobile build infrastructure.
2019-2023. Amazon Shopping. Senior Software Engineer and Tech Lead. Key work:
- Led platform programs across navigation, shared UI, and build infrastructure, coordinating across 10+ teams and 50+ stakeholders.
- Built foundational platforms that now serve hundreds of millions of users daily with zero critical incidents at scale.
- Solved a 2-year navigation challenge that had defeated previous engineering attempts.
- Led SPM migration achieving 81% conversion rate and 54% p90 build time reduction, saving approximately 6,000 engineering hours per month.
- Ran biweekly demos attended by 70+ Directors, VPs, and Distinguished Engineers.
- Influenced the 3-year platform roadmap for Amazon Shopping’s mobile experience.
- Mentored dozens of engineers across multiple teams.
2024. Shipped Buy For Me, an agentic AI shopping experience for Amazon, coordinating across 10+ teams on a 100-day timeline. The product was publicly announced on aboutamazon.com and covered by TechCrunch, The Verge, and others.
Completed MIT Sloan’s AI Business Strategy program. Left Amazon after seven years.
2025. Started building Intarsia. AI automation for SMB back-office operations, designed for non-technical operators.
The thesis: the same workflows that large enterprises automate with expensive, complex systems can be made accessible to small and medium businesses through AI agents that operate with human supervision. This is the “management by exception” pattern from the 2016 blog post and the “autonomous execution with human supervision” stage from the 2017 Intel talk, now made practical by the current generation of large language models.
Education
- MSc in International Management, Bocconi University, Milan (110/110)
- CEMS MIM, Aalto University, Helsinki
- AI Business Strategy, MIT Sloan, 2024
- Stanford Machine Learning (Andrew Ng), 2011
Languages: English, Italian, Spanish.
Location: Seattle area.