Instead of listening to music or making calls on his way to work, Mitchell Karstetter tunes into his favorite podcast. It’s not a celebrity or news pundit; it’s two digital hosts, powered by artificial intelligence, talking about data from his farm during harvest season.
“I like that it’s breaking down everything that we’re doing,” says Karstetter, the owner of RJK Farms. “It’s giving me real-time data that I can use to make decisions faster.”
Karstetter is using Innov8.ag’s HarvestReplay system to collect data and make decisions based on their workforce. Each day, it gathers data from the farm and then synthesizes and relays it to growers in the form of an audio podcast.
“With HarvestReplay, they now have access to the kind of operational insight they have gotten used to having on the row crop side of their business,” says Steve Mantle, CEO and founder of Innov8.ag.
He says right now, this new technology is helping specialty crop growers due to the labor-centric nature of the business, but there are plans for growth into other areas of farming.
How HarvestReplay Scanners Provide Real-Time Insights
For the HarvestReplay to work, they need an automated labor and tracking system. FairPick and FairTrak, which have two ways to track how much a single employee has harvested throughout the day, are examples of these systems.. The first option is for employees to put their harvested product on a scale-like scanner, where it tracks work output, such as pounds per hour picked. The second option is to have their badge scanned by a phone-like handheld scanner to report their statistics. At RJK, they are currently using it on around 600 acres of apple and cherry trees. This allows farmers to track and follow worker’s efficiency and ultimately their productivity.
“It gives you something where you can go, ‘Why is this guy, who’s normally my best guy, not performing as well,’” Karstetter explains. “It helps you identify problems faster.”
The HarvestReplay also tracks future weather conditions to help make important decisions, such as when the best times are to harvest or spray. It uses on-farm or state-operated weather sensors, such as AgWeatherNet.
It can also incorporate a grower’s harvest data from previous years to help provide insights into the farm’s historical trends. Adding it all together, AI hosts can then educate farmers on things like how early frost impacts crop volume and quality.
“It’s giving me insights into fields that were not as productive as I thought they were on cost, labor or efficiency,” says Ellie Norris, owner and CEO of Oregon’s Norris Blueberry Farms.
The Three-Tiered Power of HarvestReplay’s Data Ecosystem
The system produces three types of podcasts, depending on who’s listening and their role in the operation:
- CFO/owner podcast focuses more on economics, such as comparing orders from different buyers.
- The farm manager podcast can be either in English, Spanish and/or other languages. It discusses what happened on the farm and offers advice on decision-making for the upcoming day or season.
- The crew lead’s podcast is typically in Spanish. This revolves around recommendations for improving operational efficiencies that affect the bottom-line economics of the farm.
It’s part of a bigger smart data interface. The podcast is only one-third of the HarvestReplay system:
- Replay History looks to turn multi-year harvest and labor records into reports and goals. This shows the past performances and economics of previous harvest decisions.
- Replay Live gives same-day feedback using GPS labor tracking. It can raise or flag issues such as congestion, slowdowns or misallocated crews so managers can adjust.
- Replay Podcast is an AI-generated audio briefing built from the grower’s own harvest data.
“Farms don’t have data analysts, IT teams or CTOs,” says Mantle. “HarvestReplay handles the heavy lift of data aggregation and integration while keeping their data private and the decision-making customized to their operations.”
Leveraging Real-Time Data to Protect Farm Profitability
In 2025, there was a 46% increase in U.S. farms declaring bankruptcy from the previous year. HarvestReplay’s goal is to target areas where farms lose money such as labor, crop production and decision-making.
The system uses the data it collects to recommend changes in order to provide a path for growers to save money. Karstetter shares an example of quickly using the HarvestReplay’s information to switch things up.
“We can go into a block and see that some cherries have been on the smaller side, so we need to prune heavier,” Karstetter explains.
It also allows him and his managers to make decisions on the crews that are working. Karstetter says that in the latest podcast entry, it shared that one group was being more productive than the other. Now he can use this information to see what one group is doing differently and how it sets them apart.
“It’s giving you a kind of like a real-time look at what you’re doing and where you’re at,” Karstetter says. “We really don’t have that unless you sit down and input all this stuff manually.”
