Internship Thesis Student for Edge AI Optimization for VLMs & VLAs (fmd)

The future starts here! Ready to join NXP

Are you inspired by the rapidly evolving world of Edge AI and Efficient Generative Architectures on edge devices and excited about exploring how these technologies can make a meaningful impact in realworld applications Do you want to contribute to cutting-edge research and development that optimizes AI workloads while gaining handson experience in a supportive hightech environment

Join the Artificial Intelligence Competence Center at NXP and play a key part in shaping the future of intelligent systems.

Your Team

At the AI Competence Center we welcome motivated students who are enthusiastic about Edge AI model and system optimization methods and their applications in emerging edgeborn use cases such as Edge Agentic AI. You will join a collaborative team focused on research innovation and engineering in Edge AI.

You will have the opportunity to explore and develop stateoftheart model compression methods and inferencetime optimizations for Small Language Models and Vision Language Models. Together with the team you will evaluate the performance of optimized models and systems on NXPs embedded systems and will help potential integration of developed methods in NXPs embedded systems.

Your responsibilities

• Exploring designing and implementing model compression techniques (quantization sparsity knowledge distillation etc.) inference optimizations (e.g. speculative decoding) and efficient generative architecture design.
• Documenting findings with clarity and supporting internal knowledge sharing.
• Working closely with team members to identify promising directions for future development.
• Communicating research outcomes through scientific publications and/or invention disclosures.

Your Profile

• Currently pursuing a Masters degree in Computer Science Artificial Intelligence Machine Learning or a related field.
• Great analytical and problemsolving skills with the ability to work both independently and collaboratively.
• Excellent English communication skills for interacting with a diverse multinational team across multiple sites.
• Curious openminded and eager to explore new technologies while contributing to meaningful and impactful AI research.

Preferred Skills

• Very good understanding of AI/ML concepts (including LLMs VLMs Agentic AI) and experience with frameworks such as PyTorch and TensorFlow.
• Familiarity with model compression techniques such as quantization pruning and knowledge distillation.
• Understanding of LLM inference optimization techniques including speculative decoding.
• Experience with Python and modern software development practices (modular design testing).
• Basic knowledge of Linux and Git.
• Experience with edge AI deployment (e.g. TFLite ONNX ExecuTorch) is a plus.

Duration

This is a full-time internship (36/40 hours per week) with a duration of minimum six months or longer. Please note that to be considered for an internship/working student assignment at NXP you need to be registered as a student during the entire period of the assignment.

We encourage applicants to upload their grade transcripts as part of their application. While optional it helps us gain a clearer understanding of your academic background when reviewing your profile.

More information about NXP in the Netherlands...