R & D Research Intern

R&D Intern – Radar Sensor Data Calibration and CloudProcessing Optimization

About Agora NZ

We run a live, automated calibration engine for radar-monitored milk vats across New Zealand. Each vat's calibration is continuously refined in the cloud from real-world data, with no manual tuning.

It works. Now we want it ready for a much bigger network, and that's where you come in. This is a real production system with real engineering discipline behind it, not a toy project.

At Agora, we believe smarter farming starts with better data. Agora grew out of Deosan's two decades of dairy innovation. It started as an internal project and is now a national platform: secure, wireless telemetry from milk vats and across whole farms. No hard wiring. No manual uploads. Just reliable data built for real-world conditions.

We're a small team, so your work gets seen and used.

The challenge

A dairy shed is a hard place for radar. Steam, foam, agitators and stainless steel all shape the signal, and every vat is a slightly different shape. Our engine handles that by learning each vat's calibration from reference data over time, with layers of checks built in.

The next problems are interesting ones:

• Scale. Calibration is compute-heavy. As the network grows, we want processing that scalesefficiently, without growing the bill.

• Day-one calibration. A brand-new install has no collection history yet. How good can we make its first calibration, as early as possible?

• Trust signals. Which statistical signals tell us, early and cheaply, how much confidence to place in a sensor's data?

What you'll do

• Get under the hood of the existing pipeline: Python, time-series data, scheduled cloud jobs, dashboards and IoT platform integrations.
• Profile where the time and memory go, then design and prototype a faster architecture: incremental processing (refit only what's changed), parallel runs, event-driven calibration triggered by a new collection, and lighter reporting.
• Research and test better methods for calibrating new installs, using real field data as your testbench.
• Build data-quality and sensor-health signals that strengthen the whole network.
• Ship your work properly: tested, reviewed, deployed through git, and documented so the nextperson can pick it up.

What's in it for you

• A real production codebase.
• Staged rollouts, automated monitoring, failure alerting and carefuldesign decisions. You'll learn how serious data systems are built and run.
• Real data at scale.
• Months of real telemetry from working farms across the country.
• Impact you can measure.
• Speed, compute cost and calibration quality are all measurable.You'll see the before and after.
• Proper mentoring.
• A supervisor at Agora, plus day-to-day support from the engineers behindour data platform.
• Farm time. Get out on site and see the sensors behind the data.

Who we're looking for

Currently studying towards, or graduated within the last 12 months with, a Degree, Master's orPhD in Computer Science, Software Engineering, Data Science or a related field.

Strong interest in cloud architecture, data processing algorithms and IoT sensor networks.

A hands-on problem solver who wants to build reliable backends connected to real-worldinfrastructure.

Good with Python and comfortable with maths: regression, least squares and some geometry.Bonus points for Linux, cloud platforms, time-series databases, dashboards or performanceprofiling.

Must be legally permitted to work in New Zealand.