Introduction to Sakana AI Applied Team for Job Candidates
Sakana AI ● Covered by 3 sources
Sakana AI is growing its Applied Team, the group turning its AI research into real products for finance and defense clients. Founded in 2023, the startup wants Japan to matter again in a field long dominated by the US and China.
Based on reporting by Sakana AI — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Sakana AI spun up an Applied Team in early 2025 to take its generative AI research out of the lab and into the hands of paying customers. The company, founded in 2023, says partnerships with major domestic and international players are already underway, with projects aimed at reshaping how entire industries operate.
The near-term focus is deliberately narrow: finance and defense, two sectors the company describes as foundational to Japanese society. Rather than spreading thin across every possible market, Sakana AI is betting that a startup's limited resources go further when aimed at problems with outsized impact. People from well-known Japanese tech firms and veteran professionals from various industries have been joining the Applied Team, and the group is expanding quickly.
The founding story matters here. One co-founder spent time at Goldman Sachs, including a stint running rates trading for the firm's Japan operation before moving into AI research at Google. Another co-authored the original Transformer paper in 2017 before later transferring to Google's Japan division. A third built a career across Japan's foreign ministry, the World Bank, and an executive role at Mercari. Their pitch: AI development has clustered around the US and China for years, and Japan, as a major democracy in Asia, should have a real seat at that table rather than watching from the sidelines.
Inside the company, the Applied Team works alongside a Research Team of more than thirty researchers, roughly seventy percent of whom are from outside Japan. The two groups aren't formally tied to each other. Business projects don't dictate research topics, and applied engineers aren't required to use whatever the researchers happen to be building. But the ideas cross over anyway, often informally, through weekly randomized lunches the company calls Fish Lunch, plus reading groups and study sessions. Techniques from an automation project called AI Scientist, for instance, have already found their way into applied work.
Sakana AI also keeps publishing papers and releasing code publicly, a choice framed as both philosophical and practical. The company argues global AI progress depends on shared knowledge, and that a Japanese firm contributing serious open research is part of what keeps Japan relevant in the field at all. On hiring, the company says plenty of current employees came in without prior AI experience, relying instead on self-directed learning and genuine curiosity about where the technology is heading.
My take — AI-written commentary, not fact-checked reporting
Betting a startup's limited resources on finance and defense instead of chasing every industry at once is the kind of focus most AI companies talk about and few actually practice. The bigger story, though, is the open research commitment sitting next to enterprise contracts with banks and government ministries — that's an unusual combination, and it only works if the company keeps publishing work good enough to justify the tension. Hiring people without AI backgrounds because they've clearly done the reading on their own time is also a quiet rebuke to an industry obsessed with pedigree.
Read more about this at: Sakana AI
