About Ovum

Built by people who know what a missed heat costs.

A team from precision livestock tech and computational biology, building per-cow breeding analytics for Australian dairy farmers who need reliable heat detection, not guesswork.

Mission

What we are here to do

Lift conception rates across Australian dairy by making heat detection as reliable as soil temperature readings.

Not a vague platform for "transforming agriculture." A specific tool that tells each farmer which cow to inseminate, and when, with a confidence score attached. The data exists in the collars already on your herd. We read it per-cow, not per-herd, and surface it as an actionable alert before the window closes.

Ovum was founded in Melbourne in 2023 by Ariella Heffernan-Marks and Tom Brandvik. We are independently funded and building for farmers, not for a trade sale.

The Team

Four people, one shared frustration with herd averages

Ariella Heffernan-Marks

CEO & Co-Founder

Former product lead at a dairy herd management software company in Gippsland. Spent six years watching farmers make breeding decisions on population averages that their own herds didn't fit.

Tom Brandvik

CTO & Co-Founder

PhD in computational ecology from the University of Melbourne, specialising in time-series behavioural models for livestock. Built the first version of the Ovum prediction model as a research prototype in 2022.

Priya Subramaniam

Head of Dairy Science

Veterinarian with a background in bovine reproductive physiology. Joined Ovum to ensure the model's predictions align with what actually happens biologically. She interprets the edge cases the algorithm flags.

Callum O'Shea

Customer Success

Grew up on a dairy farm in the Western District. Helps new farms through the collar connection process and the calibration period. The first call farmers make when something doesn't look right.

Our Story

Why we built Ovum

Tom came from computational ecology. He had spent four years building time-series behavioural models for wildlife tracking before shifting his focus to livestock in 2021. The core technical problem was the same: extract signal from noisy activity data, account for individual variation, and predict a future state.

Ariella had been on the other side of the problem. Working in dairy herd management software, she had spent six years watching farmers use alert systems that fired based on herd-average thresholds. A cow that naturally runs high activity was being flagged constantly. A quiet cow that was genuinely in heat was being missed. The tools didn't know the difference because they had never been asked to learn it.

The two met at an AgTech meetup in Melbourne in late 2022. Ariella described the false-positive problem. Tom described the individual-baseline approach he had been using for wildlife behaviour. The connection was immediate.

The first version was a Python script running on Tom's laptop, ingesting CSV exports from one farm's collars. By April 2023, it was a working prototype predicting breeding windows with 75% confidence on that farm. By July 2023, Ovum was incorporated and running on three farms. The prediction engine has been refined continuously since then, with Priya joining in early 2024 to ensure the biological accuracy of every threshold and confidence score the model produces.

Our Approach

Three principles we don't compromise on

Per-cow, not per-herd averages

Every alert Ovum issues is specific to one cow's history and one cow's current activity pattern. We do not apply herd-level thresholds to individual animals. This is the foundational design decision. Everything else follows from it.

Works with existing collars

Ovum is a software layer. It reads the data your existing collars already produce. We will not build our own collar hardware, and we will not require you to replace equipment that works. If a collar produces standard 3-axis accelerometer data, Ovum can ingest it.

Honest reporting of confidence

Every alert carries a confidence score. If the model is uncertain about a cow (irregular cycle, early calibration period, or recent health disruption), we say so. You see the number. You make the call.

Independently funded. Building for farmers, not investors.