Data Scientist
This role focuses on adapting computer vision models for resource-constrained edge devices in Grab's Geo Vision team, applying compression techniques like quantization and pruning to maintain accuracy. You will profile on-device performance and build efficient inference pipelines to support map creation and localization services.
Skills required
About this role
The Data Science (Geo Vision) team improves Grab's maps and builds map-based intelligence (localization, routing, travel-time, traffic forecasting) using computer vision and deep learning on images, video, and GPS signals.
As a Data Scientist, you'll focus on bringing computer vision models to embedded/edge platforms for map creation. You will be working onsite and reporting to the Head of Data Science based in Cluj.
What you'll work on
- Help adapt and improve CV models (detection, segmentation, tracking) to run on resource-constrained edge devices.
- Apply model compression — quantization, pruning, knowledge distillation — to shrink models while preserving accuracy.
- Profile on-device performance (latency, memory, power) and help build efficient inference pipelines.
- Support training and evaluation, and use AI agents/tooling within the team's quality guidelines.
What you'll gain: hands-on experience taking deep learning models to on-device deployment, plus skills in model optimization and real-time edge performance.
Qualifications
- Pursuing or recently completed a degree in Computer Science, Electrical Engineering, or a related field.
- Familiarity with Python and some of: PyTorch/TensorFlow, NumPy, OpenCV, and scikit-learn.
- Solid ML and deep learning / computer vision fundamentals.
- Interest in or exposure to embedded systems, edge deployment, or model optimization (quantization, pruning, distillation) — a strong plus.
- Proficiency in English
Additional Information
Our Commitment
We are committed to building diverse teams and creating an inclusive workplace that enables all Grabbers to perform at their best, regardless of nationality, ethnicity, religion, age, gender identity sexual orientation, and other attributes that make each Grabber unique.
