About the job
# About the Team
- The AI Platform team is on a mission to create a platform that allows anyone to utilize AI technology quickly and reliably, providing technical support for AI utilization across Toss.
- The AI industry is rapidly evolving from LLM calls to autonomous agent systems that plan, utilize tools, and execute tasks independently.
- We are working at the forefront of this transition in two key directions.
- Supporting Service Teams: We design and operate a platform that enables various service teams at Toss to quickly experiment with agents and deploy them reliably into production.
- Enhancing Internal Productivity: We create and disseminate an internal AI automation environment, allowing all Toss members to leverage AI to reduce repetitive tasks and enhance efficiency.
- We are looking for someone who enjoys defining and structuring technically ambiguous problems.
# Responsibilities
- Collaborate directly with various internal teams to identify bottlenecks and translate needs into technical requirements, linking them to the platform roadmap. Quickly validate onsite requirements through PoCs and contribute the results as platform features.
- Build an integration layer connecting internal systems (messengers, data pipelines, work tools) with LLMs, creating an environment where anyone can utilize AI-based automation.
- Design and operate the execution infrastructure that enables agents to plan, call tools, and perform long-running tasks reliably.
- Design a memory system that allows agents to remember and utilize previous conversations, user preferences, and work history. Structure short-term (session-level) and long-term (user/team-level) memory, managing what information to store and when to retrieve it at the platform level.
- Design common infrastructure and developer experiences that enable individual teams to create and operate agents tailored to their domains.
# Ideal Candidate
- Experience applying technologies such as LLMs and agents to real service challenges is highly desirable.
- We prefer individuals who have defined unstructured problems technically and solved them systematically.
- Experience in defining and understanding problems in complex and dynamic situations, and connecting them to technical solutions is a plus.
- We want candidates who have collaborated with multiple teams to develop and operate technology as a product.
- A proactive approach to following the fast-paced advancements in AI technology and incorporating them naturally within teams is preferred.
- Interest in simplifying complex AI systems into consistent and user-friendly experiences is advantageous.
# Preferred Qualifications
- Experience designing and operating agent systems requiring multi-step reasoning or multi-agent collaboration is a plus.
- Experience designing or managing long-running agent workflows that connect planning, execution, tool invocation, and state management is desirable.
- Context engineering experience that involves designing how and when to provide information for agents to operate correctly is preferred.
- Experience designing and operating memory systems at conversation, session, and user levels is advantageous.
- Experience designing or operating quantitative assessment systems to evaluate agent quality is a plus.
- Experience directly communicating with internal teams to uncover needs and translate them into technical specifications for the platform is desirable.
- Experience automating repetitive tasks or internal processes with AI, leading to everyday usage by team members is a plus.
# Resume Recommendations
- Start by explaining the problems you have solved rather than focusing on the tech stack, and detail why you chose specific structures.
- Describe how you ensured operational reliability and scalability from experimentation to production in detail.
- Include experiences on how you detected and resolved issues such as failures, performance degradation, or cost problems.
- Document the process and results (usage rates, feedback, etc.) of uncovering internal team needs and connecting them to platform functionalities.
- If applicable, include any open-source contributions.
# Joining Toss
- Application submission > Job interview > Cultural fit interview > Reference check > Compensation negotiation > Final acceptance.
- The job interview will focus on in-depth technical discussions and ML system design.
# A Note for Potential Colleagues
- “We are not stopping at merely calling LLMs; we are building systems where agents autonomously plan, utilize tools, and achieve long-term goals. From experimentation to deployment — we design the entire flow at Toss.”

