Wifi valid intern

Location: Sophia Antipolis, France (FR)

Experience Level: Intern

Description

About Ceva

Ceva is at the forefront of the Smart Edge revolution, with innovative state-of-the-art Silicon and Software solutions that enable products to Connect, Sense and Infer. 

Within the Wireless Internet of Things (WIoT) Business Unit, we are offering you a unique opportunity to shape the future of connected devices. Our advanced wireless technologies, including Bluetooth, Wi-Fi UWB, and 5G Cellular IoT, are integrated into over a Billion devices annually, driving innovation across industries. 

About the job

A connected product doesn’t behave the same way depending on its orientation, its channel or its distance to the access point. To check that, our measurements literally turn: the product sits on a motorized turntable sweeping through the angles while we measure throughput, latency and range.

That’s daily life for our WiFi Validation team: performance, regression and compliance campaigns making sure every feature of our IPs delivers on its promise. Today, those campaigns still rely heavily on manual work — orient, restart, collect, consolidate — time the team could spend on higher value-added topics.

In practical terms, the internship will be organized into several phases:

Control the measurement testbed

Automatically control the bench equipment: turntable, device under test, commercial access points, programmable attenuators, and traffic generators

Check the state of the bench before each campaign, detect faulty equipment and bring it back to working condition automatically, so a campaign can run overnight unattended

Chain the campaigns

Automatically chain rotation tests: at every turntable orientation, replay the throughput, latency and range scenarios across the configurations to cover (channels, bands, data rates, link options)

Log, alongside the measurements, the radio link indicators and the test conditions that make campaigns comparable with one another

Repeat measurements to characterize their spread and build a reliable basis for comparison

From a single interface, select the equipment, the products and the tests to run, order the campaigns and launch them without touching the testbed

Make the results talk

Structure results in a common format, indexed by version, product and configuration, and connect them to our validation chain

Compare a campaign against previous versions and automatically flag regressions, telling a real degradation apart from ordinary measurement spread

Surface the tests that ran, the measurements and their trends in a dashboard accessible to the whole team

Speed up investigation (AI component)

Automatically group similar failures and gather the useful context — logs, configuration under test, campaign history — so no one starts an analysis from scratch

Link a performance drop to the indicators logged at the same instant to propose a first hypothesis on the cause, opening the door to machine learning on the accumulated history

Requirements

• Final-year engineering school or Master's degree student.
• Writing scripts (Python) and have some grounding in web development or databases
• Proficiency working under Linux
• Curious about technology and how things work

• Duration: 6 months
• Start date: Early 2027