Hiring: Agent Engineer

The role focuses on creating and evolving our AI agent infrastructure, including agent architectures, tool integration frameworks, reasoning systems, and an all-automation mindset. This infrastructure serves both our customer applications and core agent capabilities.

The agent systems share in common the need for flexible reasoning and action capabilities. AI agent services also require care for scalable decision-making both in frontend interactions and backend processing, as well as advanced monitoring and behavior logging. Interest in these issues should make this role attractive, as we are working at creating a great work environment to tackle them.

The role has a lot of freedom in technology choices, which we expect to decide as a team. The needs for flexibility and intelligence lead us to consider, for now:

• LLM orchestration platforms (prompt engineering, chain-of-thought, ReAct patterns)
• Autonomous agent frameworks with tool use capabilities
• Memory systems for context retention and reasoning
• Evaluation frameworks for agent performance and safety

The role is also about joining a team dedicated to the future of AI, to work with research scientists, ML engineers and designers. The agent systems support all these activities on a daily basis, and benefit from our work on advanced automation. This complementarity is at the core of the team and the company business model, to allow us progress toward an advanced AI world and provide the best AI tools on the market.

Please consider contacting us to apply! We expect a few things from you:

In terms of technologies and skills, we are interested in experience with:

• Agent development frameworks (LangChain, AutoGPT, BabyAGI, etc.)
• LLM integration and prompt engineering
• Tool integration systems for agents
• Planning and reasoning systems
• Agent evaluation methodologies
• Fine-tuning and RLHF techniques
• API development, e.g. RESTful and protobuf
• A programming language, e.g. Python, Rust, Go, JavaScript
• Distributed version control, basically Git

We have no experience level requirement, as we wish to grow together with you and, well, there is no need for everything right away. So please consider applying even if you only see partial coverage of the above!

If you score yourself high on all the items, we definitely hope for a chance to talk with you. We will consider some extras:

• Experience with “real world” agent applications and concrete AI solutions
• Multidisciplinary background or interests, notably related to cognitive science, linguistics, psychology, and ethics
• A real passion for AI, beyond LLMs
• Graduate degree in related disciplines
  • Degree is not necessary to apply, but...
  • Degree is sometimes necessary to apply for visas
  • The role would benefit from advanced analytical skills

We hope applying to work with us is a great experience. We value your time and your interest! We think the best way to decide working together is first to know more of each other, starting with us sharing our online profiles.

We propose meeting a number of times, primarily online, and freely talk about you, about us, and about what we could achieve together. As much as we would like to meet every applicant, we expect first reading about your motivation for applying, and seeing your past achievements. This should give us a chance to reply early whether we should meet for the occasion. We promise to reply to each serious application.

In short, here is a typical application flow:

1. You apply, sending us a resume and a cover letter. Again, we value very much seeing your public contributions.
2. We invite you for a first online meeting, or explain why we may not meet this time./li>
3. We may organize other meetings, for you and us to know better, sometimes on site.
4. We expect to make decisions relatively rapidly.
5. We may ask for references.

How we work together is in flux, as the team and company grow. We aim at adapting working conditions as we all see fit. Currently:

• We encourage "truth seeking"
• Personal growth-oriented environment
• Honest salaries
• Hideaway office within 1 min walking distance from Takadanobaba station
• Some prefer remote work, some office work, some change every day
• We gather at the office when needed
• We breathe open source, when it makes sense
• Inclusive
• Medical checkup support in Tokyo
• We love challenges and opportunities that make us evolve

We hope these items paint a constructive image of our work style. All these items depend on the team, and your presence will influence them, as a partner.