Sakana AI represents a fundamental shift in the artificial intelligence landscape. Based in Tokyo, this research lab is not simply another entity chasing the scaling laws of large language models. Instead, it is pioneering a nature-inspired approach to collective intelligence. For professionals seeking a career that balances cutting-edge research with real-world implementation, Sakana AI offers a unique environment characterized by its "breeding" of AI models and a commitment to sustainable technology.

To apply for a position at Sakana AI, candidates must follow a specific dual-step protocol. First, complete the official Google Form available on the company's career portal. Second, send a brief self-description and a confirmation of the form submission to careers@sakana.ai. It is critical to ensure that all official recruitment communications originate only from the @sakana.ai domain to avoid potential scams.

The Scientific Mission Driving Sakana AI

Understanding the mission of Sakana AI is essential for any prospective applicant. Unlike many Silicon Valley startups focused on brute-force scaling—using more data and more compute—Sakana AI looks to the natural world for inspiration. The name "Sakana" (Japanese for "fish") reflects this philosophy: the power of a school of fish acting as a single, intelligent entity.

The lab focuses on developing efficient AI technology through methods like evolutionary computation and model merging. Instead of training models from scratch, they "breed" them by combining existing open-source and closed models using tree search and evolutionary algorithms. This approach is not just a technical preference; it is a strategic response to the resource constraints of our current era.

The vision expanded significantly with projects like "The AI Scientist," a multi-agent system designed to automate the entire scientific research process, from hypothesis generation to writing peer-reviewed papers. Joining Sakana AI means participating in a paradigm shift where AI is not just a tool but an autonomous engine for discovery.

Decoding the Core Research and Engineering Roles

Sakana AI is expanding rapidly following its successful Series B funding round, which raised 20 billion yen. This capital is being deployed to strengthen both their foundational research and their "Applied Team."

Member of Technical Staff (MTS)

The Member of Technical Staff is the cornerstone role for researchers and engineers. This position is ideal for individuals passionate about pushing the boundaries of generative AI, open-endedness, and autonomous agents. The team values candidates who have a track record of publications in top-tier venues such as NeurIPS, ICLR, ICML, and JMLR.

However, Sakana AI differs from traditional academia. While publications are valued, the ability to write high-quality code and contribute to open-source projects is equally vital. The lab looks for "full-stack researchers" who can move seamlessly from a conceptual whiteboard discussion to implementing a complex evolutionary merge in Python.

Applied Research Engineer

The Applied Team was launched to bridge the gap between frontier research and societal implementation. Applied Research Engineers work on translating Sakana AI's unique technologies into solutions for Japan’s most critical sectors, including finance, defense, and manufacturing.

This role requires a unique blend of technical depth and domain understanding. You might find yourself building multi-agent systems for a major financial institution one day and developing secure AI solutions for public sector infrastructure the next. This team prioritizes candidates with experience in RAG (Retrieval-Augmented Generation), function calling, and the practical deployment of LLM agents in high-stakes environments.

Software and Platform Engineering

Building a "research-first" company requires a robust infrastructure. Software Engineers at Sakana AI handle everything from frontend development using React and TypeScript to backend systems in Go or Python. The lab also recruits Principal Platform Engineers to manage the complex orchestration of compute resources across cloud platforms like AWS and GCP, often utilizing Infrastructure as Code (IaC) tools like Terraform.

A key highlight for engineering roles is the active use of AI-assisted coding tools. The team integrates tools like Claude Code, GitHub Copilot, and Cursor into their daily workflow, reflecting their belief that AI should enhance human productivity within the lab itself.

What Sakana AI Looks for in a Candidate

Beyond technical skills, Sakana AI hires for a specific mindset. The company operates at the "messy" intersection of experimental research and product deployment.

Technical Judgment over Rote Coding

During the interview process, you will likely find fewer standard LeetCode-style algorithm questions and more deep-dive conversations about your technical choices. The interviewers want to know why you chose a specific architecture, how you identified its limitations, and how you navigated ambiguity during a project. They value technical judgment—the ability to assess risks and prototype effectively to test those risks early.

Communication and Critical Thinking

Because Sakana AI collaborates with large enterprises and government entities, the ability to communicate technical concepts to non-technical stakeholders is paramount. This is especially true for roles like Technical Program Managers and Project Managers. The team seeks individuals who can distill complex problem spaces into actionable roadmaps.

The Innovation Mindset

The lab follows a guiding principle: "Only do the research that wouldn't happen if you weren't working on it." They are not interested in incremental improvements to existing Transformer architectures. They want bold thinkers who are willing to explore "strange" and speculative ideas, such as moving beyond the Transformer toward new architectures like the Continuous Thought Machine (CTM).

Does Sakana AI Require Japanese Language Proficiency?

This is one of the most common questions from international applicants. The answer depends heavily on the specific role.

  • Research Roles (MTS): Business-level Japanese is generally not a requirement. The research environment is highly international, and English is the primary language for technical collaboration. However, candidates are expected to relocate to Tokyo.
  • Applied and Business Roles: For roles that involve client interaction, such as Product Managers, Applied Research Engineers, and Communications Managers, business-level Japanese proficiency (typically JLPT N1 level) is often required. This is because these roles involve co-creating solutions with Japanese institutions where the working language is Japanese.
  • Bilingual Advantage: Regardless of the role, being bilingual in English and Japanese is a significant asset. Sakana AI aims to democratize AI in Japan while maintaining a global research standard, making bridge-builders highly valuable.

Navigating the Application and Interview Process

The recruitment process at Sakana AI is designed to be a conversation rather than an exam. It aims to reveal how a candidate thinks and works in real-world scenarios.

Step 1: The Initial Application

Ensure your CV and cover letter are in English. Even for roles requiring Japanese, the initial review process often involves international stakeholders. In your cover letter, be explicit about the role you are applying for and why Sakana AI's specific mission resonates with you.

Step 2: The Take-Home Project

For engineering and research roles, Sakana AI often utilizes take-home projects. Unlike timed coding tests, these projects are meant to be completed over a few days. The evaluators are looking for:

  • Depth of Reasoning: Documentation that explains your thought process.
  • Code Quality: Clean, maintainable, and well-structured code.
  • Handling Ambiguity: How you make decisions when the requirements are not 100% defined.

Step 3: Technical Interviews

The interviews are typically treated as technical discussions. You might be asked to present a past project you are particularly proud of. Prepare to defend your design choices and discuss what you would have done differently with more resources or time.

Step 4: Cultural and Final Rounds

The final rounds often involve speaking with the leadership team. These conversations focus on long-term alignment with the company's vision for sustainable and sovereign AI.

Relocating to Tokyo and Office Culture

Sakana AI is firmly committed to being a Tokyo-based company. They believe that Japan, with its unique challenges like a declining workforce and aging population, is the perfect "living lab" for implementing AI that truly benefits society.

Visa and Relocation Support

For international hires, Sakana AI provides comprehensive assistance with the visa application process. While the primary base is Tokyo, the office culture typically follows a hybrid model, combining in-office collaboration with remote flexibility.

The "Startup" Energy

With a team size of 51 to 200 employees, Sakana AI maintains the fast-moving energy of a startup. The team is roughly half Japanese and half international, creating a multicultural environment that avoids the "Galapagos" effect often seen in isolated domestic markets.

The Future Growth and Opportunities

The Series B funding has set the stage for Sakana AI's next phase: deployment. This transition from pure R&D to business and public sector implementation creates diverse opportunities beyond pure research.

Focus on Sovereign AI

There is a growing global demand for "Sovereign AI"—models tailored to specific national cultures and values. Sakana AI is focusing its R&D on the post-training layer to optimize frontier models for Japan's specific needs. This creates a unique niche for professionals interested in the intersection of AI policy, culture, and technology.

Expansion into New Sectors

The company is scaling its partnerships with organizations like Mitsubishi UFJ Financial Group (MUFG) and Daiwa Securities. They are also expanding into defense and manufacturing. This diversification means they are hiring for specialized roles, such as Project Managers with backgrounds in defense or manufacturing analysts who understand how AI can transform industrial workflows.

Frequently Asked Questions

What is the best way to get noticed by Sakana AI recruiters? Submit a high-quality GitHub repository or a list of peer-reviewed publications. The team values tangible proof of your ability to build or discover.

Does Sakana AI offer internships? Yes. Sakana AI offers various contract types, including 4-month internships, part-time student researcher roles for graduate students in Japan, and 1-year visiting researcher positions.

Where is the Sakana AI office located? The office is located in central Tokyo, Japan. Most roles require candidates to be physically based in Japan or willing to relocate.

Are there non-technical roles available? Yes. The company is hiring for business operations, accountants, recruiters, and internal auditors to support its rapid growth.

What is the "AI Scientist" project? It is a multi-agent system developed by Sakana AI that automates scientific discovery. It can generate ideas, code experiments, and write full research papers, representing a core part of their research into autonomous intelligence.

Summary of the Sakana AI Career Path

Sakana AI is a destination for those who find the current AI "arms race" unsustainable and are looking for a more intelligent, nature-inspired alternative. Whether you are a researcher aiming to move beyond the Transformer or an engineer looking to implement AI in the heart of Japan's financial and defense sectors, the lab provides a platform for high-impact work. The combination of a world-class founding team, significant financial backing, and a mission focused on "AI for Japan and the world" makes it one of the most compelling places to work in the global AI ecosystem today. To join, focus on demonstrating deep technical judgment, a willingness to tackle ambiguous problems, and a commitment to the sustainable evolution of intelligence.