#170 【FinTech】LLM Research Scientist

※This job posting is for applicants residing in Japan.

Company Overview

Our client provides Revenue-Based Financing (RBF), a financing method in which future projected revenue is purchased as receivables.
Using proprietary algorithms to forecast future sales, they operate a platform that enables fast and easy fundraising without equity dilution or collateral.
They have established Japan’s first RBF fund and completed approximately JPY 1.4 billion in total funding (Series A), and continue to grow rapidly.

Their team consists of members with diverse global backgrounds, and Over 40% of their members are foreign nationals.They offer the opportunity to work in a multinational team (Asia, Europe, Central America, etc.).
They also offer a full-flex time system, hybrid work (generally 1–3 days in the office per week depending on the role), and full remote options depending on conditions.

Job Description

You will participate in a research and development project for a next-generation privacy-preserving AI platform. This is a groundbreaking challenge in a world-first technological field, integrating large-scale language models (LLMs) and cryptographic technology.

This project aims to fundamentally solve the challenges of AI utilization in regulated industries (finance, healthcare, law, etc.), and plans to submit papers to top international conferences and release open-source software over a two-year research and development period.

・Research and development of large-scale language models (architecture design, implementation, evaluation)
・Design and implementation of new algorithms
・Benchmark evaluation and performance measurement
・Paper submission and peer review for international conferences
・Writing technical blogs and documentation
・Technical sharing and mentoring within the team

【CTO Profile】

Master’s Degree, Graduate School of Information Science and Technology, The University of Tokyo
Physics Olympiad Silver Medal (2012, 2013)
International Conference Acceptance: SIGGRAPH Asia 2019
Ethereum Foundation Grant Acceptance
ETHGlobal Tokyo 2023 Finalist
Devcon SEA 2024 Speaker
Specialization: Cryptography, Secure Computation, Distributed Systems

Required Qualifications

• Master’s degree or higher (in a related field such as computer science, information engineering, mathematics, or physics)
• Deep understanding of and implementation experience with Transformer models
• Experience implementing large-scale models using frameworks such as PyTorch/TensorFlow
• Ability to write and read academic papers in English
• Advanced programming skills in Python
• Japanese ability: Not required

Preferred Qualifications

• PhD (Computer Science, Information Engineering, Mathematics, Physics, etc.)
• Experience in building large-scale language model learning and inference infrastructure
• Paper acceptance record at top conferences (NeurIPS/ICML/ICLR/ACL, etc.)
• Practical experience in Fine-Tuning, RLHF, Prompt Engineering, etc.
• Knowledge of cutting-edge technologies such as Agentic AI, RAG, and Context Management
• Experience in implementing distributed learning and inference
• Proven track record of contributing to open-source software
• Interest in cryptography and security technologies

Compensation & Benefits

Salary: Negotiable depending on experience and skills (includes 45 hours of fixed overtime per month)
Stock Options: Granted after joining (based on performance)
Bonus: None
Salary Increase: Once per year
Probation Period: 6 months
Location: Tokyo, Remote Work (2 days remote, 3 days in office per week)
Social Insurance: Fully covered
Working Hours: Super flex (no core time; 8 working hours + 1 hour break between 5:00–22:00)
Holidays: Weekends, public holidays, and New Year holidays (Dec 30 – Jan 3)
Average Overtime: 15 hours
Smoking Policy: No smoking indoors

Selection Process

Selection process: 3 interviews + aptitude test + reference check
Documents: Resume, CV
Interview Language: Japanese or English