Agriculture is a uniquely challenging industry when it comes to lending, thanks to the wide variety of information needed to understand a farm’s bankability. Numbers about land parcels, livestock health, equipment, crop pricing, and production history all have to be taken into account, and today, most ag lenders are ill-equipped to do this efficiently.
Technology can obviously support that and help with that, but it hasn’t been smart enough to figure out exactly what should be collected from a farmer in every given situation,” says Luke Johnson, founder and CEO at agrifintech company SweetAg.
Founded as Landjourney in 2024, the company has developed an agrifintech platform that supplements the often-cumbersome legacy processes often used in agricultural lending.
“It’s important to make the process of ag lending more efficient, and to allow a farmer to know exactly what needs to be done to get a loan, securely,” says Johnson. “Technology can help that.”
SweetAg recently raised $7.4 million in a funding round led by Diagram Ventures, Builders VC, and Cooperative Ventures. The new capital will go towards growing the Boulder, Colorado-based company’s client base.
AgFunderNews (AFN): What problem in agrifinance does SweetAg solve?
Luke Johnson (LJ): Legacy platforms in banking are very rigid, with lots of compliance rules. But those platforms were built in an era when today’s AI tools did not exist. As a result, they have not been able to absorb the utility of AI tools natively—it’s very, very hard for modern AI tools to interact with banking platforms in an organic way.
Meanwhile, farm banking is defined by customized categories that don’t really sit in any area of a loan origination system. There’s all this operational data, collateral information, legal and financial documentation to gather on these farms to understand if they’re well run and have financeable operations. Today’s ag lenders have limited access to tools that provide the interactivity necessary to gather all of this directly from farmers.
[The industry needs] a way to incorporate modern interactivity and AI to make it so that farmers and their lenders don’t have to go through some exotic journey just to get information into their system and approve a loan.
AFN: Why is it so difficult for farmers to access capital right now?
LJ: Fifty years ago, a farmer had a relationship with a local banker, and that was the best way to access capital. Today, banks don’t even have the ability to issue a loan approval based solely on a relationship. All loans produced must be fully documented, and supported by viable operations.
Of course, relationships are important in this industry, great lenders spend a lot of time with farmers helping provide incredibly valuable advice, and both parties benefit from meaningful relationships. But the point is that loan approvals are not based on friendships, no matter how strong the relationship is, and technology becomes meaningful to support even the most traditional relationships with efficient ways to document, approve, price, and fund a loan.
AFN: What is unique about SweetAg in relation to all of this?
LJ: There isn’t a lot of fintech that is actually core banking infrastructure targeting this sector. There are a lot of point solutions, and strong point solutions at that, but there isn’t any core banking infrastructure that is built to target this segment at all. This is partly because core banking infrastructure is very difficult to execute and you really have to know what you’re doing to build it correctly.
At SweetAg, we have a unique team that’s built core banking loan origination systems and lending infrastructure, and we know the space, soup to nuts. We are the only fintech targeting core banking systems in all of ag.
The industry has a lot of very specific needs from a fintech perspective that are not being met. There are clear gaps in their modernity and their access to really fundamentally basic tools that we expect and take for granted everywhere else.
Farmers have migrated to online banking because rural bank branches have all but evaporated, payments are efficient online, and almost every bank has fairly good access to online checking. Even so, most aren’t yet able to be fully serviced by most agriculture lenders online. Now, agricultural lenders need to modernize in order to, for example, allow an applicant to apply and fully document a loan online, supply regular information for covenant checks, or renew a loan.
We originally created SweetAg to fill this gap. Ag lending includes information on land parcels, the health of livestock, long equipment lists, crop production histories, crop prices, and a team of people (including extended family) supporting farm operations. We build technology specifically for the segment to meet the industry where it stands, rather than providing some rigid commercial banking platform a lender has to invest a ton of time customizing to force it to accept the data or process that a lender follows to support ag lending.
It’s important to make the process of ag lending more efficient, and to allow a farmer to know exactly what needs to be done to get a loan, securely. Most ag lenders in the country work via email with farmers, which is really non-compliant and extraordinarily dangerous from a security perspective. Email chains feel never-ending and never give the team supporting a farm (or lending organizations) the full picture in one place is needed to obtain a loan.
Technology can obviously support that and help with that, but it hasn’t been smart enough to figure out exactly what should be collected from a farmer in every given situation, and that’s one of the problems we solve in this space.
Some of the best existing technology that could support this information collection from farmers was also built to collect consumer information from one individual at a time. This, again, wasn’t built for agriculture or commercial loan production, and doesn’t work when some things need to be collected from a spouse, a brother co-borrower, or a CPA. In addition to working fluidly with the entire supporting team, we make the process smart, make it know exactly what it needs no matter what the structure is of the loan, the farm, and what is being produced.
AFN: How does AI specifically address such issues?
LJ: What’s interesting about this industry is that you’re getting a lot of different formats of information and a lot of different styles of documentation that supports a lending decision. And when that happens, AI can be uniquely useful in applying a bank’s quality control checks and the ability to extract data natively from those documents in an efficient way.
For example, if a farmer uploads a balance sheet that’s hand written, tools today can be useful for absorbing that information and understanding whether it lines up. If financials don’t foot, a farmer can be alerted immediately, well prior to weeks of back and forth with an underwriting team.
That allows the industry to operate like it does natively in its own format in its own processes, with a farmer still uploading, say, a balance sheet that is handwritten on a piece of paper, and still be able to get through a lending process efficiently.
AFN: SweetAg emphasizes the need for humans in this process along with AI. Elaborate on that.
LJ: AI is really good at boring stuff. I’ve run a lending shop myself at pretty large scale, and I can say that continuously nagging your team to check every document — for example, ensuring the name and address on the statement matches on every document — becomes a futile and sad exercise.
Machines, on the other hand, are very good at doing the same boring thing consistently, like checking to see if a tax return has every single page in it, or if it’s dated.
Machines are also really great at reminding humans to do their part. No lender wants to burn relationship points pestering a farmer to supply a document for a covenant, but farmers might find automated reminders helpful and less intrusive.
Where human judgment comes into play is when, for example, there’s an inconsistent number: someone entered one number in an application and there’s another number in a source document. That happens all the time in banking, and there are many reasons a number could be justifiably different. It’s a lender’s job to work through those different numbers, and at this point, humans are much better at evaluating and arbitrating that difference.
Along the same lines, there are a lot of reasons why humans are very good at making discretionary decisions and judgements about whether to loan money to a farm at a certain rate. The idea is not to get humans out of the loop, it’s to make consistent and efficient all of the human activities that are most annoying, unreliable, inconsistent, or tediously slow.
It is also important to recognize when algorithms should be applied instead of AI. Algorithms are very, very good at, for example, applying consistent ratios or consistent eligibility rules. That should be a deterministic algorithmic equation, not just AI saying “I think that this should apply.”
AFN: What about data privacy concerns among farmers and their lenders?
LJ: Any time there’s a move from paper to online, there will be concerns with security and privacy. What are you going to do with this data? Are you sending it to Anthropic or ChatGPT? How do I make sure my data is secure?
Refusing to use AI is like refusing to acknowledge the internet. The cat is out of the bag, and internal limits can backfire spectacularly. We can ring-fence data with our system, but when a rogue employee uses a personal ChatGPT account to pull data from a set of financials because their executive team fears AI, all of the security and controls that could have been systematically put in place go right out the window.
It is worth stating explicitly that all of this is better than what is being done today, because basically most commercial lenders are operating with paper and email in a completely unsecure environment today – our tools work to encrypt, secure, and make access trails auditable.
However, because this transition is new for many, and a loan officer or borrower might not be accustomed to it, there is some education that’s necessary to assure them this is indeed more safe and secure than email or paper.
The good news is that this is a well-worn path because all of these farmers are already banking online. It is absolutely an expectation that lenders operate online with some modicum of modernity. By definition, agriculture is the audience the internet was actually made to serve, because they are broadly distributed and it’s not always convenient and easy for rural farmers to access a physical banking location.
