Software Engineer

About Recursive

As the world's population grows, we are facing a number of global challenges in the realms of environmental conservation, societal equity and ensuring that nobody is left behind in the transition to a digital economy. Artificial intelligence is a key technology to not only address some symptoms of these issues, but also helping create completely new innovative approaches that can create a more sustainable economy and lifestyle.

Recursive is an Applied AI consulting company based in Japan working on developing solutions to some of these challenging problems. We collaborate with large Japanese enterprises to create innovative solutions to sustainability challenges, merging in our expertise in AI research and our client's domain knowledge. Our development projects focus around the main themes of improving efficiency, accelerating innovation, risk mitigation and better work and education.

Recursive is founded by Tiago Ramalho and Katsutoshi Yamada in 2020. Tiago was a Senior Research Engineer at Google Deepmind and Lead Research scientist at Cogent labs and brings years of machine learning expertise to the company. Katsutoshi is a serial entrepreneur and AI expert with experience in developing and establishing innovative business models in Japan and abroad.

Building on its previous successes, Recursive is now looking to grow and expand its impact further. Visit our homepage to learn more about Recursive.

 

Recursive’s Mission

At Recursive, our mission is to attract individuals who share our passion for our core purpose: to create a fairer, more sustainable society through the power of Artificial Intelligence.

Our Mission:

• Why: To build a fairer, more sustainable society.
• How: By transforming our clients into AI-first businesses.
• What: Through AI technology development and consulting.

In pursuit of this mission, Recursive has distinguished itself in several key areas:

• Cutting-Edge AI Application: We excel in translating state-of-the-art AI research into practical applications. This is achieved through strong connections with academic machine learning research, a commitment to understanding client data and requirements, and seamless integration with client businesses.
• Generative AI Innovation: We launched a Generative AI solution to the market just as the field was gaining momentum, having initiated development in early 2022.
• Enterprise AI Solutions: We introduced FindFlow only six months after ChatGPT's release, enhancing the application of large language models for enterprise use and enabling access to private enterprise data with Generative AI.
• Pioneering Environmental AI: Recursive is the first Japanese company to employ Physics-Inspired Neural Networks for simulating environmental systems. We aim to use these methods to transform environmental management and promote renewable energy adoption. Learn more about this project by watching our presentation at the Google Cloud for Sustainability event.

 

Recursive’s Culture

Ultimately, the value of a company is the result of the collective actions of its passionate members. At Recursive we want to focus on creating an incredible working environment where talented individuals can come together to work hard on solving real-world problems.

When it comes to culture, we know that there is no way to suit everyone’s tastes. Almost everyone has a passion, but not all passions align. And the same is true for Recursive.

We aim to build a work environment that is ideal for a particular kind of person:

• Challenger Mentality: Those who like to solve hard problems and try to go beyond their current abilities
• Leadership: those who take the initiative and fix problems on their own
• Trust & Cooperation: those who seek to work with a team of talented individuals they can trust
• Impact: those who want to feel like their work is having impact on others/the world Recursive can only work when we take collaboration to be our top priority.

Our workplace has zero tolerance for harassment or toxic behavior, deliberate bad performance, as well as any behaviors that undermine the culture we wish to create. We will only be able to achieve the goals of closing more impactful projects, and achieving “kandou” for the ones we are executing, if all teams come together and collaborate.

We want to use our diversity of cultures and experiences as a strength instead of a weakness. Given that we are still a small company, everyone’s actions matter, and we place top priority in ensuring that our culture (i.e. the actions of our members) reflects our values.

That means that it is not enough to simply write down a document with what we believe our culture is, but in a nutshell, actively rewarding behaviors that are consistent with the culture and disincentivizing behaviors that go against the culture that we want to build.

Recursive’s Employment Framework

・Contract Structure: Recursive provides one year fixed term contracts for all new employees. The contracts are converted into permanent contracts upon new members meeting Recursive standards and contributing effectively to Recursive’s success. 

・Flexible Work Hours: Recursive operates on a flex system with the core hours of 10:00 - 16:00. The flexible work hours fosters a supportive environment that accommodates diverse employee needs while ensuring effective team collaboration during core hours.

・Hybrid Work: Experience the best of both worlds with our hybrid work model, offering up to three days of remote work and two days in our Shibuya office.

・Comprehensive Leave Policy: Enjoy public holidays, winter break, 20 days of paid annual leave, 5 days of sick leave along with special leave for bereavement and paid parental care.

・Enriching Welfare Programs: Take advantage of our diverse welfare programs such as health insurance, annual health checkups, employee’s welfare pension, accident and unemployment insurance, paid parental leave, caregiving leave, babysitter voucher program, team-building allowances, skill development and language learning support, relocation assistance, and visa application support.

・Supportive Benefits: Access a range of benefits including commuting allowance, referral bonus, employee invention profit share program, and stock option plan for eligible roles.

・Open to Side Projects: While side jobs are generally not allowed, we are open to discussions based on individual circumstances, fostering a supportive and flexible work culture. Some of our members teach at universities, contribute to non-profit organizations, or engage in pet projects that improve their knowledge and skillsets.

 

Responsibilities

• Prompt & Pipeline Engineering: Design and implement robust prompt-tuning and orchestration pipelines using frameworks like LangChain or LlamaIndex. This includes techniques such as prompt chaining, few-shot prompting, and developing autonomous agents.
• Model Integration & Deployment: Integrate and deploy LLMs (both proprietary and open-source) into production environments. This involves managing API integrations, handling request/response schemas, and ensuring models are performant and cost-effective.
• Data & Knowledge Systems: Design and build data pipelines for Retrieval-Augmented Generation (RAG) systems. This includes creating vector stores, optimizing embedding strategies, and ensuring data is up-to-date and accessible for real-time model queries.
• Backend & API Development: Develop and maintain high-performance, fault-tolerant backend services and APIs that serve as the core of our AI-centered applications.
• Performance Benchmarking: Conduct quantitative analysis and A/B testing to benchmark the performance, latency, and cost of different LLM prompts and models, and iterate on solutions based on data-driven insights.
• Cloud Infrastructure: Utilize cloud platforms (GCP, AWS, Azure) to deploy, manage, and scale applications and their supporting infrastructure.
• Collaborative Development: Work closely with frontend developers and product managers to integrate LLM-powered features seamlessly into the user experience and deliver value-driven solutions.

 

Expectations

• Action-Oriented Problem-Solving: Possess strong analytical and debugging skills to rapidly diagnose issues and implement solutions in a fast-paced, iterative development cycle.
• Production-Focused Mindset: A history of delivering on-time, fit-for-purpose, and robust software that provides tangible value.
• Communication & Teamwork: Ability to work effectively within cross-functional teams, communicate complex technical concepts clearly, and give and receive constructive feedback.
• Proactive Learning: Stay current with the latest advancements in LLMs, open-source tools, and cloud technologies, and be willing to safely experiment with new solutions to improve our products.

 

Qualifications

• Education and Experience: Hold a Bachelor’s degree, or 10 years of industry experience in a software development role
• Programming Languages: Experience in at least one systems language (Python, Go, Java, C++, Rust, etc.)
• Database Experience: Experience with database management and operations, including relational and NoSQL databases, and an understanding of vector databases.
• Cloud Proficiency: Proficiency with setting up, deploying, operating, and maintaining applications with at least one major cloud provider, such as AWS, GCP, or Azure
• Applied LLM Experience: Demonstrable experience in building and deploying at least one application powered by a large language model. This could include a personal project, a contribution to an open-source tool, or professional work.
• Benchmarking and Performance: Familiarity with benchmarking and quantitative analysis methods to assess and optimize AI and application performance
• Passion for the Field: Passionate about AI and how it applies in the present, and vision for how it will apply in the new future.
• Language Ability: Proficient in English, able to discuss in depth technical topics in it; familiarity with Japanese is a plus.

 

Interview Process

We aim to make the hiring process smooth and transparent. The typical interview steps are as follows:

• 1. First Interview : With Hiring team
• 2. Pair Programming Interview
• 3. Third Interview : With Engineering team
• 4. Final Interview : With Leadership