Harvesting says its agri intelligence engine will not only make it easier for banks and microfinance institutions to offer agri loans and monitor farmlands, but also double farmers’ income in the long run.
At a glance
Founder: Ruchit Garg
Founded in: 2016
Based out of: Fremont, California (India office in Bengaluru)
Services: AI and data-powered credit solutions for agri-lenders
Sector: Fintech (focused on agricultural finance)
Funding raised: $400,000 (including a grant from Catalyst Fund)
In the middle of 2016, Wired magazine ran a piece titled The future of humanity’s food supply is in hands of AI. It spoke of “smart farming”, aided by emerging technologies like artificial intelligence, machine learning and data analytics. “Machine learning algorithms may help usher in a new Green Revolution to keep humans fed on an increasingly mercurial planet,” it said.
What if a regular farmer could be notified of weather inconsistencies through predictive analytics, such that his harvest is unaffected? What if farmlands could be monitored remotely using satellite technology? What if agri-lenders could take stock of lands, farm loans, credit records, repayment cycles etc., on a real-time tech-enabled platform?
Harvesting, a California-based startup, enables all of the above and more. Founded in 2016 by Microsoft alumnus and serial entrepreneur Ruchit Garg, who underscores that he “grew up in India” and is aware of its agrarian challenges, Harvesting is about to begin its operations here in the next few weeks.
It has set up an office in Bengaluru and is onboarding talent across departments, from designers and data scientists to product and sales managers. The firm says it is in talks with all leading private banks of India and some PSUs too for a potential business opportunity.
What Harvesting does
Harvesting operates at the intersection of fintech and agri tech, and its product is powered by AI, deep learning, and data analytics.
It has built an agri intelligence engine that utilises remote sensing and geo-spatial data alongside myriad traditional and alternative data points to assess a farmer’s creditworthiness. It then shares this knowledge with banks and microfinance institutions (that dole out farm loans), in the form of a credit scorecard.
It also offers region-specific farmland data to these agri-lenders and financiers, and helps expedite the process of lending by easing access to information and by reducing middlemen. Whenever on-farm performance varies from predicted performance beyond a certain threshold, the lending institution receives automatic notifications.
On the farmers’ side, the AI-powered engine conducts remote monitoring of farmlands, captures changes in vegetation or crop cover, and provides early warning systems such that harvesting risks and damages are minimised. This not only helps reduce wastage in the agricultural value chain, but also makes farmers more efficient, capable and smart, with greater access to financial services and better loan repayment records.
“There are over 500 million small farmholders in emerging markets that feed 80 percent of the world. But there is a data asymmetry in the agricultural value chain. Most problems arise because of a massive data deficit between farmers and financial institutions. We started to look at how this could be solved by leveraging data and technology.”
At Harvesting, over 11,000 computers equipped with deep learning and AI techniques process satellite images from NASA, European Space Agency and the likes and turn them into data insights, projections and credit scorecards. The startup has roped in highly qualified geo-scientists to pull this off. “The process requires a combination of human and artificial intelligence,” says Ruchit.
Enabling agricultural finance
Harvesting says banks deal with tens of thousands of farmers regularly. Thus, it is humanly impossible for them to stay on top of each record in real time. It leads to inaccuracies, delays and even turning down of loans. Sometimes, lenders fail at loan recovery and collection too, because either the farmer is incapable of repaying or the bank has lost sight of the farmer or his land.
“We use data-driven insights to help BFSIs and MFIs monitor lending and loan utilisation. Our product helps agri-lenders look at the farmers’ financial scorecard. It also provides them crop data directly from the farm. If a loan recovery is due, we notify them 15 days in advance. It reduces the turnaround time significantly.”
This not only enables banks to dole out more “affordable products” for farmers, but has the potential to have a long-term impact on agricultural finance. Ruchit believes it could even bring down interest rates and double the income for farmers.
As it happened with coffee farmers in Uganda, where Harvesting executed a state-funded project, and helped bridge the deficit between farm and finance. Ruchit reveals his firm is now engaged in conversations with the World Bank too.
While Harvesting is bringing its entire product suite to India, it is specifically targeting to solve the problem of land records.
“Most banks in India mortgage a farmer’s land when giving him a loan. Keeping track of that becomes a big problem later on. Managing those documents, understanding the language nuances etc., are challenges.”
Hence, Harvesting is roping in patwaris (a government official who maintains land records) in every region or state to solve this. Furthermore, land documents are usually written in the local language(s), which computers cannot accurately process. Harvesting is now building a Natural Language Processing (NLP) capability to find a solution to this uniquely Indian problem.
Harvesting says it is having conversations with 60-plus global financial institutions for the deployment of its product. So far, its work has been concentrated in East and West Africa, parts of Latin America, and Southeast Asia. India, which is a major agrarian economy, is a market tailored for disruption next.
Harvesting also expects to close its first VC funding this year. So far, it has run on a grant from Catalyst Fund, some dollars from an angel investor, and an initial investment from a “prominent Indian family” (name undisclosed).
“We wanted to build something impactful that could change the lives of people. We have managed to build a scalable business by leveraging satellite imaging technology AI and can help the government double the farmers’ income.”