Kento Sasaki
Career Background
During university, I interned at the National Institute for Materials Science (NIMS), working on materials informatics research and conducting a project to automatically extract experimental settings and compound information from materials science papers using NLP techniques. I then joined Turing as an intern in June 2022, contributing to data infrastructure development and ML model development for the Gen-1 data collection vehicle.
Why I Joined Turing
Around June 2023, Turing had just opened a new office at the Kashiwa-no-ha Campus. At that time, co-founder Aoki reached out to me via Twitter DM to invite me to intern. I think the team had at most one member beyond the co-founders at the time.
As an intern, I initially worked in a small two-person team with Yu Yamaguchi, who is now CTO. While overwhelmed by the talent around me, I strongly felt my own growth through working with Yamaguchi. Drawn to this environment, I decided to join as a full-time employee.
What I Do
In-Vehicle Infotainment Development
After joining full-time, I first worked on IVI (In-Vehicle Infotainment) development, mainly handling vehicle CAN signal analysis and HAL (Hardware Abstraction Layer) implementation. AOSP (Android Open Source Project) is an extremely large-scale project with numerous APIs and services across layers, and I remember struggling to grasp the full picture. During my first year, Team Manager Shotaro Watanabe taught me the fundamentals of software development, for which I remain grateful.
GENIAC Project Phases 1–3
About a year after starting IVI development, I joined the GENIAC project and have worked across Phases 1, 2, and 3. I have been involved in various R&D efforts including the VLM evaluation benchmark Heron-Bench (CVPR 2024 WS), the autonomous driving VLA model training dataset and baseline model CoVLA-Dataset (WACV 2025), and the image compression tokenizer One-D-Piece (ICML 2025 WS). Most recently, I built STRIDE-QA (AAAI 2026), a dataset for spatiotemporal understanding in driving scenes.
VLA Team Initiatives
Currently, as VLA team leader, I am advancing R&D on VLA models that look beyond End-to-End autonomous driving models.
What Makes My Work Rewarding
Exceptionally talented people from diverse fields create an environment where you can always work with a drive to improve. The in-office-first policy enables smooth face-to-face communication, making it easier to get things done.
There are many opportunities to serve as a core member of projects and teams, and I have been fortunate to attend international conferences including CVPR 2024 (Seattle), WACV 2025 (Tucson), ICCV 2025 (Honolulu), and AAAI 2026 (Singapore).
About the Work Environment at Turing
- With professionals in every domain, you can quickly consult whenever you face difficulties.
- Systems and structures are in place to maximize development efficiency.
Who Thrives at Turing
- Takes ownership and works proactively
- Aspires to take on significant work with a high-level perspective
- Interested in both engineering and research (for the VLA team)
For Those Considering Turing
At Turing, every engineer is given decision-making opportunities, and everyone has a chance to make an impact. It is an ideal environment for those seeking growth as young engineers and researchers, and for those who want to work on R&D with real business impact.
Fully autonomous driving is a challenge in a domain humanity has not yet reached. We face non-trivial problems every day, and we would love to work with people who can enjoy tackling such challenges.
My Car
Suzuki Jimny (5th Gen)
I chose the Jimny because I wanted a tough car for easy trips to the beach or mountains. Delivery is scheduled for March. I currently drive a Toyota Yaris Cross, and since the TRC office (Heiwajima) allows car commuting, having my own car makes the commute more enjoyable.
Cars I’d Like to Drive
I have always admired 2-door coupes and am interested in the Audi TT. I also thought the Honda Prelude I saw at Japan Mobility Show 2025 was really cool.
