We don't wait for potential to show. We predict it.
Every herd pays to raise replacements that will never earn their keep. HerdGen reads the weights you already record during the growing period and tells you which young animals are worth keeping — months before the usual answer. No lab tests, no new equipment.
What late culling costs your farm
Move the sliders to your own figures. The estimate counts the rearing cost sunk into replacement females that leave the herd before they have paid for themselves, and the share of those that early selection can identify while they are still growing.
Conservative by design: the model counts only rearing cost avoided, not the extra offspring, milk yield and weaning weight recorded in the original trial.
HerdGen EarlySelect
You enter the weights. The platform computes growth-intensity indices and growth-model parameters for every animal, ranks the group and tells you which to keep. Thresholds are calibrated per species, breed and country. Below is the working group view.
Replacement females, 3rd weighing
48 animals · Large White × Duroc · updated today| ID | Growth | g/day | Score | Decision |
|---|---|---|---|---|
| A-117 | 468 | 94 | keep | |
| A-104 | 441 | 88 | keep | |
| A-129 | 418 | 79 | keep | |
| A-142 | 404 | 61 | watch | |
| A-133 | 371 | 38 | cull | |
| A-108 | 342 | 24 | cull |
Illustration of the group view in development. Pigs are the first species implemented; others follow the same structure.
Four steps, using data you already collect
Weigh three times while they grow
Individual live weights at fixed ages during the growing period. That is the only input the method needs.
Model each animal's growth
Growth-intensity indices and the Bridges and Richards growth models describe how an animal is forming, not just how heavy she is today. They apply to any species with a measurable growth curve.
Rank against calibrated thresholds
Thresholds are fitted to your species, breed and conditions. In pigs, growth energy above 410 g/day predicts a litter above 10 at first farrowing.
Keep the right ones
A clear keep, watch or cull list months before first breeding, and a mating plan for the animals you keep.
Proven first in pigs, by design
A method is only worth as much as the trial behind it. Ours began with a full three-factor experiment conducted in 2008–2011 on a breeding farm in southern Ukraine: 120 replacement females, 48 dams, 192 animals on feeding trial and 48 carcass evaluations, all weighed and recorded by the author. The dissertation was defended in 2016 before a specialised academic council, with the head of Ukraine's Institute of Pig Breeding among the official opponents.
Pigs were chosen because they give the fastest generation turnover and the harshest test of a prediction: an answer at six months can be checked against a real litter at twelve. The underlying relationships — growth intensity in early life and later reproductive performance — are not specific to one species, which is why the same structure is now being prepared for sheep, goats and cattle.
And on working farms, not only in the trial
From 2016 to 2025 the selection criteria ran on commercial farms in southern Ukraine under long-term service contracts. Two of them became reference farms: production herds where every group of replacement females was evaluated the same way, year after year, and where the approach proved itself in farm economics rather than in an experiment.
Both owners invested years of disciplined weighing and record-keeping in this work. Much of what HerdGen EarlySelect automates today was first done by hand, group by group, on their farms. The same approach has since been applied on other farms in Ukraine and Armenia.
Methodological guide (2026): doi:10.5281/zenodo.22745011. Results were implemented on the trial farm and adopted in a university curriculum in animal breeding.
- Litter size of purebred dams was significantly determined by growth energy during the growing period.
- Animals selected by the thresholds produced 1.7 more offspring per litter, 12.1 kg more milk yield and 52 kg more litter weight at weaning.
- Crossbred females from lighter groups showed compensatory growth: litter weight at weaning rose to 160.7 kg against 115 kg in controls.
- Thresholds established for pigs: growth energy above 410 g per day, litter size above 10 at first farrowing.
Science sees what others have to wait for.
Dr. Artur Ishkhanyan, founderBuilt for farms of any size
Large integrators can afford genomic testing and their own geneticists. Small and mid-size farms cannot, and they are the majority of producers in the United States. HerdGen needs nothing but a scale and the records you already keep.
- Enter weights from the growing period, or import from your records
- Indices and growth-model parameters computed per animal
- Ranking against species-, breed- and country-specific thresholds
- Keep, watch or cull list with the prediction behind each score
- Track how last year's decisions played out in the herd
What we can do for your herd now
Herd assessment
Send your weight records for a group of young females and get them back ranked.
- Keep, watch or cull list
- Thresholds fitted to your species and breed
- Written explanation for each decision
Selection consulting
Ongoing work on the structure of your breeding herd.
- Weighing schedule and recording
- Mating plans and breeding method
- Replacement and rotation planning
Research collaboration
Joint work extending the method to new species, breeds and climates.
- Joint trials and publications
- Guest lectures and student projects
- Partners in the US and Ukraine
Dr. Artur Ishkhanyan
Animal breeding and genetics scientist and founder of HerdGen Analytics. Doctorate in Animal Sciences, specialising in breeding and selection of livestock. He spent three years on the farm collecting the data behind this method himself, then a decade applying it in commercial herds. Before moving to Los Angeles he ran a wholesale agricultural market in southern Ukraine that was selected as a USAID AgroInvest pilot project, and worked at the Askania-Nova Institute of Animal Husbandry.
Questions farmers ask
Do I need genomic testing?
No. The method works on live weights recorded during the growing period. If you already weigh your replacement stock, you have everything the model needs. Genomic data can be added later, but it is not required.
Does it work for my species and breed?
Pigs are implemented and validated. For sheep, goats and cattle the same models apply, but thresholds must be fitted to your own records first — that is the first step with any new farm, and we are open about it.
How accurate is the prediction?
In the trial, breeding method and live weight during the growing period explained up to 22.3% of the variance in reproductive traits and 31% of litter weight at weaning. That is a meaningful signal for a culling decision, not a guarantee for an individual animal, which is why borderline animals are flagged as "watch" rather than culled.
What does it cost?
A herd assessment is priced per group; consulting is a monthly retainer. The platform will be offered as a subscription scaled to herd size. Write to us with your species and herd size and we will send figures.
When will the platform be available?
It is in development, with pilot farms in California planned for 2027. Until then the same analysis is done as a service, and pilot participants get the platform first.
Talk to us
Tell us about your herd, your species and how you weigh. We reply within two business days, in English, Ukrainian or Russian.
- Company
- HerdGen Analytics LLC, California
- Office
- 15233 Ventura Blvd, Suite 500, Sherman Oaks, CA 91403
- Support
- support@herdgenanalytics.com
- Research
- Kherson State Agrarian and Economic University, Ukraine