When Emrick Donadei started his career at Google, working on artificial intelligence was not even on his radar. Similar to numerous engineers at the tech giant a few years back, his responsibilities were completely separate from the realms of large language models (LLMs) or AI safety. The 32-year-old, now based in New York, openly admitted he felt unqualified for an AI position at the time.
The ChatGPT Catalyst and the Internal Shift
The landscape within Google began to transform dramatically with the launch of OpenAI's ChatGPT. This event acted as a catalyst, pushing the company to accelerate its own work on large language models. As Google opened up internal pathways for engineers to switch teams, Donadei's interest was piqued, though he was uncertain about his place in this new AI-driven environment.
"Fundamentally, it's similar to my last role," Donadei explained. "But instead of building software, I'm building LLMs, which requires data, training, and compute." Despite the conceptual parallels, he had no direct hands-on experience with AI products, making initial conversations with specialized AI teams a challenge.
The Hackathon Breakthrough and Proactive Networking
Donadei acknowledges that the high demand for AI talent played a role, but he emphasizes that personal initiative was equally critical. His pivotal moment came when he decided to participate in an internal seven-day hackathon announced by Google in 2024. He saw it as a prime opportunity to engage with "the hottest topic in the industry."
During the event, he focused on creating a small prototype. "I didn't do anything revolutionary," he stated. "I built a small prototype that wasn't super useful, but it was a good way to get started." The process forced him to grapple with less-glamorous but essential tools like LLM infrastructure, agent-based workflows, and model fine-tuning.
Instead of stopping there, Donadei strategically leveraged this experience. He actively presented his work to technical leads across Google in short meetings, using these discussions to understand their teams' goals and expand his internal network. This proactive approach helped him clearly articulate how his skills could align with Google's broader AI objectives.
Sustaining Momentum Through Continuous Learning
Donadei's commitment extended beyond a single event. In 2025, he participated in another hackathon, a project that later contributed to a public technical disclosure from Google, solidifying his standing in the AI domain. To continuously build his expertise outside of formal projects, he adopted a rigorous self-learning regimen.
He routinely uses AI tools like Claude Code for reading documentation, Gemini and ChatGPT Deep Research for case studies, and NotebookLM for managing large information sets. Additionally, he supplements his knowledge by watching educational content from experts like Andrej Karpathy and co-hosts a podcast focused on software engineering and AI.
Reflecting on his journey, Donadei doesn't attribute his successful transition to mere luck. "By granting me unlimited access to frontier technologies and a direct line to key decision-makers," he said, "the hackathon proved that I wasn't late to the AI revolution." His story underscores a viable path for professionals aiming to pivot into AI, highlighting the power of seizing internal opportunities, dedicated self-education, and persistent networking.
