Application Engineer on (Tiny) Machine Learning

INNOITALY SRL is a start-up founded in October 2023 that focuses on designing, developing, planning, and producing energetically autonomous and batteryless sensory devices within a circular economy, including data processing, both embedded on the sensor with AI and on the cloud and blockchain.

For the new office located in ROVERETO (Trento), Innoitaly is looking for an APPLICATION ENGINEER PROFILE ON MACHINE LEARNING on embedded systems with the goal to design and implement a novel framework to support the development of green and even fully sustainable Tiny Machine Learning applications distributed in an edge-to-cloud continuum.

The objective is to design and develop techniques for energy-efficient Tiny Machine Learning on low-cost and low-power microcontrollers characterized by memory-constraints and powered with an unregular supply coming from energy harvesters.

Tasks include

• efficient data collection, AI model generation (neural networks), compression, and code generation for memory-constrained IoT devices
• intermittent execution and acceleration of the trained ML model on microcontrollers.

The activity will be intersecting with the following topics

• Energy Efficient and Low-Power Computing
• Machine Learning on Microcontrollers (Tiny Machine Learning)
• Embedded Systems and Software Support
• Architectural and Hardware Support.

Requirements

• A Master degree (or equivalent) in computer science and engineering, electronics, electrical engineering or telecomunication
• English fluency
• Machine Learning Algorithms and Deep Neural Networks
• Compilers, Efficient Code Generation
• Good programming skills (Python/C/C++/Assembly Language)
• Programming Microcontrollers, and Interfacing Sensors
• (Optionally) Basic Electronics, PCBs, Circuits

Details

• Full time
• Possibility of two days of remote work
• Permanent contract

Work location: Rovereto (Italy - Trento)

Informativa sul trattamento dei dati alla pagina https://www.agenzialavoro.tn.it/Informativa-Privacy.