ABRI: AI-Driven Genetic Insights for Breeders
Understanding the Problem
Breeders often faced challenges in selecting optimal mating pairs to achieve desired genetic traits in their herds. Traditional methods relied heavily on historical data and breeder expertise, which were time-intensive and lacked precision.
Key Issues:
- Manual, intuition-based breeding decisions led to inconsistent outcomes.
- Historical data was not always reliable in predicting desirable traits.
- Time-consuming analysis of genetic performance across multiple generations.
- Limited tools to visualise and compare mating outcomes effectively.
Defining the UX & Product Vision
Objective: Develop a data-driven, AI-powered prototype that enables breeders to predict genetic outcomes for calves, helping them select the most optimal pairings based on specific breeding goals.
The platform needed to:
✔ Leverage AI & genetic datasets to provide accurate predictions of breeding values.
✔ Offer an intuitive, user-friendly interface to visualise predicted calf traits.
✔ Enable breeders to compare siresbased on historical performance, genetic markers, and AI-generated projections.
✔ Seamlessly integrate into existing breeder workflows to support decision-making.
UX Research & Prototyping Approach
In collaboration with a major Australian breed association, geneticists from UNE (University of New England), and data scientists, I led a series of workshops and user interviews to understand breeder workflows, pain points, and decision-making criteria.
Research Methods Used:
- User interviews with breeders to identify decision-making pain points.
- Field studies & shadowing to observe current selection processes.
- Heuristic evaluation of existing tools used for breeding analysis.
- Co-design workshops with geneticists & breeders to align AI outputs with practical use cases.
Key Research Insights:
- Breeders valued visualisation tools that provided an intuitive way to compare mating pairs.
- Many breeders lacked confidence in AI-driven recommendations without transparent logic.
- Real-time adaptability was crucial—users wanted to adjust mating criteria dynamically.
Solution: AI-Powered Genetic Selection Dashboard
With insights gathered, I worked closely with data scientists and engineers to build an interactive prototype that:
✔ Evaluated available sires and predicted offspring breeding values based on genetic markers.
✔ Visualised projected genetic traits using a side-by-side comparison dashboard.
✔ Allowed breeders to adjust filters and explore alternative pairings dynamically.
✔ Ensured a user-friendly interface, integrating AI-driven recommendations with manual breeder expertise.




User Testing & Validation
The prototype underwent iterative usability testing with breeders and industry stakeholders. Early feedback was overwhelmingly positive, highlighting:
- Increased efficiency in decision-making—breeders could compare pairings in minutes instead of hours.
- Improved confidence in AI predictions—clear, explainable outputs helped breeders trust the recommendations.
- Scalability for future AI-driven enhancements—laying the groundwork for long-term genetic optimisation.
Results & Future Impact
While the prototype is still in development, it has already transformed how breeders evaluate genetic pairings, proving the potential of AI-assisted breeding selection.
Key Outcomes:
✔ Established a new standard for AI-driven breeding insights, aligning genetics, technology, and breeder intuition.
✔ Set the foundation for industry-wide adoption of predictive breeding analytics.
✔ Created a scalable UX framework that can integrate additional AI models and genetic data sources over time.
Next Steps
We are now refining AI models, expanding data sets, and exploring automation features to further streamline the selection process for breeders worldwide.
I can choose sires known for exceptional traits like carcass quality, calf weaning weight, maternal instinct, or disease resistance – it helps with hand-picking the genetics to make the most profitable calves.
John Calhoun
High Country Speckle Park