About the job
About the Team You Will Join
- The Machine Learning Engineer (LLM) at Toss Securities is part of the AI Tribe. You'll collaborate with various Data, Server, Frontend Engineers, Designers, and Product Owners to tackle product-level challenges.
- The goal of the AI Tribe is to simplify complex financial and securities information, delivering only the most relevant data to individuals through data and ML-based services.
- To achieve this, we experiment with a variety of ML techniques, including NLP/LLM training and operation, AI service development, and personalized recommendations, integrating these into real products.
Your Responsibilities Upon Joining
As a Machine Learning Engineer (LLM) at Toss Securities, you will take on two primary roles. If you excel in one area, you can still make significant contributions to the team.
Model Training, Serving, and Operations
- You will train models suited for text-based challenges (e.g., classification, extraction, summarization, translation).
- Enhance model quality through fine-tuning, experimental design, and performance evaluation.
- Serve models in a live service environment and ensure their stable operation.
- Monitor performance degradation and changes in data distribution, implementing improvements as needed.
Defining AI Service Problems, Data Construction, and AI Feature Design
- Reframe complex financial and securities domain issues into solvable ML/LLM problems.
- Define and construct datasets to facilitate model training.
- Generate advanced insights based on various financial data such as disclosures, news, and financial data.
- Collaborate with multiple teams to ensure AI features operate seamlessly within products.
We Are Looking For
- Individuals with experience in creating and deploying LLM or NLP models.
- Those who have taken full responsibility for the entire cycle from data definition to model operation.
- Insights into utilizing LLMs effectively and reliably are a plus.
- Ability to logically explain why specific approaches are appropriate in problem-solving.
- Interest in the securities domain and simplifying complex information through technology.
Preferred Qualifications
- Experience in developing and enhancing a single LLM/NLP model from start to finish.
- Experience analyzing performance degradation or data changes in service operations and implementing structural improvements.
- Experience improving AI features from the perspective of user experience or product metrics rather than solely model performance.
Resume Recommendations
- Clearly state the problems you aimed to solve, the methods you chose, and the resulting changes (e.g., metrics or product modifications).
- Include improvements attempted during the operational phase, not just model development.
- Be specific about your roles within LLM, NLP, or services.
- Focus on the overall structure and your key contributions. Exclude sensitive information that cannot be disclosed publicly.
The Journey to Joining Toss Securities
- Application Submission > Phone Interview > Job Interview > Cultural Fit Interview > Reference Check > Compensation Negotiation > Final Offer and Onboarding