This post was originally published on this site.
Artificial intelligence tools aimed at drought, wildfire risk and agricultural instability are often developed far from the rural communities expected to use them. Zoho Corporation is taking a different approach by putting college students from rural and rural-serving universities in California and Texas at the center of the development process.
Zoho has named five finalist teams in From The Ground Up, the company’s first AI hackathon focused on environmental challenges affecting rural communities. The student-built projects address problems ranging from water leaks and irrigation management to crop stress and land stewardship, with an emphasis on systems that can operate affordably and with limited technical infrastructure.
For small business owners, particularly farmers, agricultural businesses and companies operating in rural areas, the competition provides an early look at how smaller, purpose-built AI systems could eventually provide practical alternatives to large and expensive AI platforms.
The five finalist teams will receive mentorship from Zoho leaders and travel support before presenting prototypes and pitch decks to a panel of industry judges at Zoho’s Pleasanton, California, office on Sept. 22, 2026. The winning team will receive a $15,000 prize.
“For decades, one of Zoho’s guiding principles was the belief that world-class software can be built from anywhere, including rural communities. It’s the reason why we’ve had an office in a rural village in Tenkasi, India for over a decade, and why we continue to discover new talent in Texas’s Rio Grande Valley,” says Raju Vegesna, Zoho’s Chief Evangelist. “We have long heard vague suggestions towards the transformative potential of AI to solve global problems from Silicon Valley CEOs, but by providing tools and financial support to dedicated students from rural communities, From The Ground Up puts them into action, while continuing Zoho’s longstanding commitment to supporting the places beyond our major cities.”
AI Designed Around Rural Constraints
Zoho structured the hackathon around a central problem facing smaller organizations: advanced technology may be available, but the cost, infrastructure and technical expertise required to use it can put it out of reach.
Teams of up to three students were asked to design an AI system or a collection of lightweight, interoperable tools addressing wildfire risk, water scarcity or agricultural instability. Zoho also asked competitors to account for the conditions faced by rural communities and Indigenous tribes rather than assuming access to large technology budgets, fast internet connections or dedicated data science teams.
The competition’s design criteria included community sovereignty, appropriate technology scale, interoperability, equitable access and environmental accounting.
Zoho encouraged contestants to use smaller, purpose-trained models that can run locally when possible. The company specifically pointed to edge deployments on modest hardware, including solar-powered sensors, as preferable to cloud-dependent systems when a task does not require a large general-purpose AI model.
That approach could have implications well beyond the hackathon. Small businesses frequently face many of the same constraints: limited budgets, unreliable connectivity and little or no in-house technical expertise. AI products that can operate locally and solve a narrow business problem may therefore be more practical for some small companies than larger cloud-based systems.
Zoho also asked students to consider the environmental costs of their technology, including the computing and water resources consumed by AI itself. Teams were encouraged to favor efficient deployments and justify heavier computing requirements based on the benefits they provide.
Water Planning Without Expensive Modeling
Team NoNiMo, made up of Texas A&M University-Corpus Christi students Noah Wilborn, Misha Stegall and Mohammed Asad Khan, is developing BASIN, a locally operated AI tool focused on water resource management.
BASIN uses public precipitation data to generate and rank auditable drought “what-if” scenarios. The goal is to help planners, hydrologists and rural water providers determine which scenarios justify the expense of more detailed hydrologic modeling.
The project grew out of the team’s experience with the 2026 water crisis in Corpus Christi. According to the hackathon materials, regional water supply models rely on hydrology data dating to 2015, while repeatedly hiring hydrologists to update modeling for new emergency scenarios can exceed available community budgets.
For small businesses, the concept highlights one possible role for AI: not replacing specialized professionals, but helping organizations decide when expensive expert analysis is necessary.
Combining Local Knowledge With Environmental Data
Team Land Memory AI, consisting of Austin Community College-Round Rock students Kuralay Biehler and Lazzat Duiseshova, is addressing water management, wildfire preparedness and agricultural sustainability through a mobile and web application.
Land Memory AI is designed to combine a farmer’s or landowner’s historical records, photographs and personal observations with environmental data. The system is intended to help users make better-informed decisions without replacing their own judgment.
That distinction may matter for small agricultural businesses. Farmers and rural landowners often possess years or decades of knowledge about individual properties that may not appear in government datasets or commercial software platforms. Systems that incorporate those records could potentially make AI recommendations more relevant to conditions on a specific farm or parcel.
Satellite and Drone Data Target Irrigation
Brothers Edgar Bello Gonzalez and Eduardo Bello Gonzalez of the University of Texas Rio Grande Valley formed Team Dos Ojos to address agricultural sustainability.
Their project combines free Sentinel-2 satellite imagery with targeted drone flights to identify crop stress. Instead of presenting farmers with complex analytics, Dos Ojos is designed to deliver simple bilingual SMS recommendations showing where scarce irrigation water should be directed during drought conditions.
The students drew on personal experience with water shortages. They grew up in Mexico and experienced periods when their city had no water, affecting harvests and farmers’ livelihoods. After moving to the Rio Grande Valley, they encountered similar drought conditions.
For smaller farms, the use of free satellite data is significant. Agricultural technology can become costly when it requires proprietary imagery, extensive sensor networks or specialized software subscriptions. A system that uses free data and limits drone flights to areas that require closer inspection could potentially reduce those barriers, although the finalist project remains in the prototype stage.
Finding Water Losses in Rural Utilities
Team CoderOP, made up of Texas A&M University-Corpus Christi students Maharshi Barot and Saumya Patel, is tackling another water problem: leaks in rural utility systems.
The team’s AI tool combines nighttime water-flow logs with free satellite imagery to identify likely leak zones. It then converts that analysis into a plain-language worklist showing the estimated amount of water lost per day, the estimated cost and the location of the suspected problem.
The design specifically targets under-resourced utilities without large technical staffs. If approaches like this prove effective, they could be particularly important to small communities where conserving treated water may be less expensive than developing additional water supplies.
The project’s plain-language interface also reflects one of the broader themes of the competition: AI tools may deliver more value to small organizations when they provide clear actions rather than another complex analytics dashboard requiring specialists to interpret.
Giving Small Farmers More Irrigation Guidance
Team Irriga, composed of East Texas A&M University students Jessica Parra Serena and Thi Xuan My Tran, is developing an AI irrigation advisor for farmers in the Texas Panhandle.
The system uses existing soil, well and weather sensor data collected by the North Plains Groundwater Conservation District across eight counties. The AI advisor is intended to make irrigation recommendations available both online and offline, giving smaller farmers access to guidance without requiring expensive one-on-one consulting.
The team’s pilot focuses on Moore County, with plans for potential expansion across the Panhandle. The students developed the idea after examining the decline of the Ogallala Aquifer and noticing a gap between the volume of groundwater information already being collected and the amount of that information that reaches farmers in an immediately useful form.
That model may be relevant to small businesses outside agriculture as well. Organizations increasingly collect operational data without having the employees, software or time needed to analyze it. Lightweight AI systems capable of converting existing information into specific recommendations could help close that gap without requiring businesses to build their own data science teams.
Promising Ideas Still Need Real-World Validation
The finalist projects also illustrate important limitations small business owners should consider when evaluating AI technology. These systems are hackathon projects, not established commercial products, and the finalists still need to build prototypes and demonstrate their ideas before judges.
AI recommendations involving irrigation, drought planning, wildfire preparedness or utility infrastructure can also carry significant consequences. The competition’s emphasis on transparent systems, auditable scenarios and preserving human judgment addresses some of those concerns, but businesses evaluating similar technology would still need to examine the reliability of underlying data and determine when professional expertise remains necessary.
Connectivity, hardware costs and access to reliable local information could also affect whether a technically promising system works in practice. A model designed for rural deployment may reduce those barriers, but it does not eliminate them.
Zoho’s requirement that teams consider community ownership and data use adds another issue for businesses to watch. Farmers, landowners and rural organizations may generate valuable operational information. Technology providers that allow customers to retain more control over that data could become increasingly attractive as AI systems depend on larger quantities of business-specific information.
The public can participate in the competition by voting for one of the five finalist teams. Public voting will contribute to a portion of each team’s overall score, with the final decision made by judges on Sept. 22. Voting remains open through 9 a.m. Pacific Time that day at https://zoho.to/ftgu-vote.
While the projects are still experimental, the competition offers a different view of how AI could reach small businesses and rural communities. Instead of relying on increasingly large general-purpose models, the finalists are testing narrower systems built around specific local problems, existing data and limited resources. For small businesses trying to determine where AI can produce measurable value, that focus on solving one practical problem at a time may be one of the more important ideas to emerge from the competition.
