ML Researcher for Frontier AI project

OneSec; Fatih; On-site; TRY 300,000 - 600,000; Full time

Benefits offered by this job

Stock option package

Job summary

OneSec in Istanbul is building a 3D-Foundation model for autonomous drone navigation. We seek a talented ML engineer to contribute to model architectures, hyperparameter tuning, and GPU training infrastructure.

You will also work on dataset generation, using open-source models to create pseudo labels, synthetic images, and curate training data. Join a fast-moving, highly skilled team with state-of-the-art tools.

Qualifications

• Strong expertise in deep learning and neural networks.
• Experience with computer vision and CNNs (e.g., ResNet).
• Proficiency in PyTorch; CUDA and OpenCV is a plus.
• Experience with dataset generation and synthetic data is a plus.

Responsibilities

• Develop model architectures, tune hyperparameters, and manage GPU training infrastructure.
• Generate pseudo labels using open-source models and create synthetic data via geometrical transformations.
• Collaborate with a talented team on a frontier-level 3D foundation model for autonomous drones.

Skills

Deep learning

Computer vision CNNs

PyTorch

Tools

CUDA

OpenCV

Job description

OneSec is an international startup that has opened an R&D center in Istanbul.

We are looking for talented Machine Learning engineers to work on our 3D-Foundation model.

OneSec is an AI & Robotics company, working on the next generation of Autonomous delivery drones.

You can find more info on our website: OneSec.com

About this job

OneSec is building a 3D-Foundation model, in one of the largest machine learning projects in the robotics industry. Using a custom dataset with over 120 million images, and modern machine learning techniques, we are building a frontier-level foundation model for autonomous drone navigation.

Your job will consist of 2 main elements:

• Machine learning: Developing model architectures, tuning hyperparameters, working on GPU training infrastructure, etc.
• Dataset generation: Using open source AI models to generate pseudo labels for internet images, generating synthetic images using geometrical transformations, filtering datasets for valuable training data etc.

Requirements

• Strong expertise in deep learning
• Experience with computer vision and CNNs (resnets etc)
• Experience with Pytorch (and ideally CUDA, OpenCV, etc)
• Attractive salary & stock option package
• Work in an exceptionally talented team
• State of the art development tools & training infrastructure