Most dairy farms that contact us about Ovum are already running some form of herd management software. DairyComp 305 is common across VIC and SA operations. Plenty of farms also use spreadsheet-based systems built over years, or recording tools from their milk company. The question we get most often in onboarding calls is not "what does Ovum do" but "how does it fit with what we're already running?" That question deserves a direct, honest answer because the integration landscape for farm software is messier than the industry tends to acknowledge.
This article covers what integration actually looks like in practice, what works well, what requires manual steps, and what we are still building toward for farms running less common recording setups.
The Core Data Flow You Are Trying to Achieve
Before covering the specific connection options, it helps to be clear about what integration is actually trying to do. The goal is to have cow identification, A.I. event records, and preg check outcomes flow between Ovum and your existing system without double data entry. Data should move in both directions: Ovum needs herd list and historical event data from your records to seed individual cow models; your herd management system needs A.I. attempt dates and conception outcomes from Ovum to maintain a complete reproductive record for each animal.
Without that bidirectional sync, you end up maintaining two sets of records that diverge over time, which creates more administrative overhead than either system on its own. The specific pathway to achieve this sync depends on which software you are running.
DairyComp 305 and Compatible Platforms
DairyComp 305 is the most commonly requested integration for farms in southern VIC and the Gippsland region. Ovum supports two connection modes for DairyComp: direct export via the DC305 format CSV, and a push API for farms running DairyComp on a network-accessible server.
The CSV export path works like this: from the Ovum settings panel, you schedule a daily export in DC305 format that produces an event file with insemination dates, alert events, and conception outcomes formatted to match what DairyComp expects when you run the import utility. That file lands in a shared folder or is emailed to the farm. Your DairyComp operator imports it at their regular maintenance window. This is not a real-time sync, but for most farms the daily cadence is adequate. Reproductive events are not time-critical at the 24-hour level; the important thing is that records stay aligned before the next preg check cycle.
The API path requires DairyComp to be running on a machine that Ovum can reach, either via your local network with the Ovum base station as an intermediary, or by exposing the DairyComp instance to the Ovum cloud service. The second option involves a port configuration in your router that some farm IT setups will not accommodate easily. If your farm's internet connection is a 4G router on a static IP, this is usually straightforward. If you are on satellite or a shared rural connection, it can require a call to your internet provider. The Ovum customer success team handles these configurations and can walk through the network requirements on a call before you commit to the API path.
Generic CSV and Spreadsheet Workflows
For farms not running DairyComp, the primary integration path is through Ovum's generic CSV export. The export is configurable: you choose which fields to include and how dates are formatted, and you can schedule automatic daily or weekly exports to a nominated email address or cloud storage location. The column naming in the export follows the conventions most common in Australian dairy record templates.
This path covers farms using farm accounting software with a livestock module, farms maintaining records in Excel or Google Sheets, and farms using their milk company's web portal for recording. The limitation is that this is a one-way push from Ovum to your records. You then need to update Ovum with the preg check outcomes from your own records, either manually or via a reverse CSV import. The reverse import in Ovum accepts a simple two-column file: cow tag number plus outcome (in-calf, not-in-calf, not-served). That file can be exported from most recording systems with a basic report function.
We know this two-step process is not ideal. The development roadmap includes a more structured bidirectional CSV sync where Ovum detects changes in a watched folder and applies the updates automatically. That feature is not yet in production, which is the honest answer to what the spreadsheet integration currently requires in terms of manual steps.
Direct Collar API Connections
The other integration question we receive regularly concerns collar hardware: specifically, whether Ovum can pull data directly from the collar manufacturer's system rather than requiring a separate base station setup.
For collars that operate via a cloud relay, where the collar manufacturer's base station or milking parlour reader uploads to their own platform, Ovum can connect via the manufacturer's API to pull the raw activity data directly. This path is currently available for the major 3-axis collar systems used in Australian dairy operations. The connection is configured once in Ovum settings with your API credentials from the collar platform; Ovum then pulls the 15-minute activity intervals at scheduled intervals and processes them through the individual baseline model without any additional data collection hardware from Ovum's side.
For collars that only record locally without a cloud component, a small relay script can be installed on the PC or NAS device already connected to the collar reader on your farm network. That script exports activity data in a standard format at each milking session and syncs it to the Ovum platform. Installation takes around 20 minutes and does not require any changes to your collar system's existing setup.
What Integration Cannot Replace
There is one data input that no integration path can automate: the preg check outcome entry. Preg check results come from your vet's visit, typically six weeks post-joining close, and those results exist on a paper form or in the vet's own recording system. Until that data is transcribed into Ovum, either manually or via a CSV import, the platform's individual cow models do not know whether a given A.I. attempt resulted in conception.
That outcome data is not nice-to-have for Ovum's accuracy; it is the primary feedback signal the model uses to refine each cow's individual threshold. A cow that conceives on A.I. attempt one but whose record is not updated for three months effectively appears as a non-conception in the model, which skews the confidence calculations for her next cycle. For farms that run one major preg check session per year, the workaround is to run a quick bulk import in the week after preg check results are available. That single annual update is enough to keep the model well-calibrated across the following joining season.
A Note on Record Duplication Risk
The integration pattern that causes the most problems is partial duplication: a farm that imports herd records from DairyComp into Ovum but then continues to manually add A.I. events into both systems. After a few weeks of joining, the two systems diverge and reconciling them requires matching duplicate records by date rather than by a clean unique event ID. The way to avoid this is to decide before joining begins which system is the record of truth for A.I. event dates. Ovum generates those events from alert-to-insemination workflow; DairyComp or your spreadsheet holds the permanent breeding history. The Ovum export should push into DairyComp, not duplicate alongside it. If your A.I. technician records directly into DairyComp on-farm and that has been your workflow for years, the practical answer is to run Ovum alerts as a detection and timing tool, export the event data, and let DairyComp remain the primary record. Ovum is not trying to replace your herd management system. It is a detection and prediction layer that should feed data into it.
Getting the Integration Right Before Joining Starts
The integration setup should happen at least three to four weeks before your joining window opens. This gives enough time to confirm the data flows are working, catch any column name mismatches in the CSV format, and ensure preg check outcomes from the prior season have been imported into Ovum to seed the individual cow models. Doing the integration setup during the first week of joining, when you need the alerts to be accurate and the team is already busy, creates unnecessary pressure and usually results in at least one sync failure that takes a day to diagnose. The setup is not technically difficult, but it needs calendar time. Contact the team before your joining season and we will walk through the configuration for your specific recording setup.
