Senior Deep Learning Engineer, Autonomous Driving & LLMs

Pony.ai; Fremont (CA); On-site; USD 140,000 - 280,000; Full time

Benefits offered by this job

Health Care Plan

401k Plan

Life Insurance

Paid Time Off

Family Leave

Disability Insurance

Free Food & Snacks

Job summary

Pony.ai in Fremont, California seeks a skilled ML/AI engineer to design, train, and deploy large-scale models for autonomous driving. You will collaborate with experts to shape software architecture, evaluation metrics, and test frameworks that power self-driving systems.

Responsibilities include building foundation models from real-world data, framing complex ML problems into solvable tasks, and deploying VLMs/LLMs with real-time performance constraints.

Qualifications

• Master in Computer Science or equivalent experience.
• Solid understanding of data structures, algorithms, parallel computing and large-scale data processing.
• Experience in applied machine learning including data collection, evaluation and feature engineering.
• Expertise in C++/Python.
• Strong communication skills and team spirit.

Responsibilities

• Work with experts on software architecture and test/ regression frameworks for self-driving vehicles.
• Design and develop large-scale foundation models trained on vast real-world data.
• Frame open-ended ML problems into well-defined ML tasks; scale to data pipelines and real-time car workloads.
• Develop and deploy deep learning models, including VLMs and LLMs.
• Optimize deep learning models to run robustly under tight run-time constraints.

Skills

C++/Python

Data structures & algorithms

Applied machine learning

Communication skills

Education

Master in Computer Science

Tools

CUDA/TensorRT

Job description

Pony.ai in Fremont, California seeks a skilled ML/AI engineer to design, train, and deploy large-scale models for autonomous driving. You will collaborate with experts to shape software architecture, evaluation metrics, and test frameworks that power self-driving systems.

Responsibilities include building foundation models from real-world data, framing complex ML problems into solvable tasks, and deploying VLMs/LLMs with real-time performance constraints.