Graduation Internship - Deployment of AI Inference Accelerator Applications

About Us

Axelera AI is not your regular deep-tech startup. We are creating the next-generation AI platform to support anyone who wants to help advancing humanity and improve the world around us.

In just three years, we have raised a total of $120 million and have built a world-class team of 180+ employees (including 55+ PhDs with more than 40,000 citations), both remotely from 11 different countries and with offices in Belgium, Switzerland, Italy, the UK, headquartered at the High Tech Campus in Eindhoven, Netherlands.

We have also launched our Metis™ AI Platform, which achieves a 3-5x increase in efficiency and performance, and have visibility into a strong business pipeline exceeding $100 million.

Our unwavering commitment to innovation has firmly established us as a global industry pioneer. 

Are you up for the challenge?

Project Summary

The accelerating pace of AI innovation continues to drive demand for efficient, low-latency, and energy-aware inference solutions in real-world applications. Axelera AI has developed a proprietary inference accelerator, called Metis, and a custom SDK called Voyager, tailored for edge and embedded AI workloads. This graduation project offers a master’s student the opportunity to explore the end-to-end lifecycle of AI model development and deployment on this custom hardware platform, from application definition to a functional demonstration system.

Supervisors:

• Industry Project Supervisor: Bram Verhoef, Director of Customer Engineering & Success, Axelera AI
• Daily Supervisor: Steven Hunsche, AI Solutions Engineer, Axelera AI

Objectives:

The primary goal of the project is to explore, develop, and demonstrate a real-world AI application, optimized for deployment on Axelera AI’s inference accelerator. The student will follow a complete pipeline from concept to deployment:

1. 1. Application Scoping & Research (Weeks 1–3):
2. Investigate and propose a relevant, real-world AI use case (e.g., theft detection, PCB fault detection, gesture recognition, etc.).
3. Define performance goals and deployment constraints (latency, power, throughput, etc.).
4. 2. Model Selection & Training (Weeks 4–8):
5. Research suitable deep learning architectures for the chosen application, being aware of the power and limitations of the Metis platform and Voyager SDK.
6. Train and validate selected models using standard datasets or proprietary data if available.
7. 3. Model Optimization & Deployment (Weeks 9–14):
8. Optimize the trained model for efficient inference using our Voyager SDK (quantization, pruning, tuning of the model).
9. Integrate and deploy the model using the Voyager SDK and Metis.
10. 4. Demo System Development (Weeks 15–20):

Expected Outcomes:

• A functional AI model tailored to a real-world application, deployed on our custom inference accelerator.
• Quantitative performance evaluation of the deployment.
• A working prototype/demonstrator that showcases the AI application in an accessible format.
• Technical documentation and a final thesis report.
• Background in machine learning or deep learning (e.g., CNNs, transformers).
• Experience with Python and AI frameworks (e.g., PyTorch, TensorFlow).
• Interest in embedded systems, AI acceleration, or edge computing.
• Familiarity with C/C++/python and Linux (Ubuntu) based environments is a plus.
• Axelera AI will provide the following facilities & resources:
• Quantitative performance evaluation of the deployment.
• A working prototype/demonstrator that showcases the AI application in an accessible format.
• Technical documentation and a final thesis report.
• Background in machine learning or deep learning (e.g., CNNs, transformers).
• Experience with Python and AI frameworks (e.g., PyTorch, TensorFlow).
• Interest in embedded systems, AI acceleration, or edge computing.
• Familiarity with C/C++/python and Linux (Ubuntu) based environments is a plus.
• Axelera AI will provide the following facilities & resources:

Required Skills And Interests:

• Access to the inference accelerator hardware and development kit.
• SDKs, documentation, and support for deployment tools.
• Weekly supervisor meetings to track progress.
• Office space and an electronics lab for testing.
• Potential access to proprietary datasets if applicable.

Learning Objectives For The Student:

• Gain experience with the end-to-end development and deployment of AI applications.
• Understand the trade-offs between model performance and deployment constraints.
• Learn to interface with custom hardware accelerators and optimize for real-time use.
• Acquire project management and technical communication skills through regular reporting and demonstrations.

Location

The internship's location is in our office in Eindhoven at the High Tech Campus, with hybrid working mode (2-3 days per week in the office) possible.

This is a paid opportunity.

At Axelera AI, we wholeheartedly embrace equal opportunity and hold diversity in the highest regard. Our steadfast commitment is to cultivate a warm and inclusive environment that empowers and celebrates every member of our team. We welcome applicants from all backgrounds to join us in shaping the future of AI.