Tani Pintar

Inspiration

Every harvest season, thousands of Indonesian farmers face the same problem: they harvest successfully but still lose money.

The issue is rarely low production, it is poor timing, lack of market information, and oversupply. Farmers often harvest simultaneously, causing prices to crash below production costs. Many are forced to sell to middlemen at a loss or even throw away perfectly edible produce because they have no access to better buyers or preservation guidance.

According to Indonesia's National Development Planning Agency (Bappenas), the country loses between 23–48 million tons of food annually, resulting in an estimated Rp213–551 trillion in economic losses. We wanted to build a solution that helps prevent these losses before they happen, not after.

What it does

Tani Pintar is an AI-powered harvest decision platform that helps farmers answer three critical questions:

When should I harvest?
Where should I sell for the highest profit?
What should I do if market conditions are unfavorable?

The platform analyzes market prices, weather forecasts, logistics costs, and crop shelf life to generate simple recommendations that farmers can immediately act on.

To make the solution accessible, farmers primarily interact through WhatsApp, receiving proactive alerts about potential price drops and oversupply while also being able to request quick information through chat. A web dashboard is available for registration, farm management, and analytics.

How we built it

We built Tani Pintar around a data-driven recommendation engine that combines multiple sources of information:

Market price data
Weather forecasts
Logistics distance estimation
Commodity shelf-life information
AI-based recommendation logic

The web dashboard allows farmers to register their farms and manage harvest plans, while WhatsApp serves as the primary communication channel for notifications and quick queries, ensuring the experience remains simple for users with limited digital literacy.

Challenges we ran into

One of our biggest challenges was balancing powerful functionality with usability.

Our primary users are smallholder farmers, many of whom are not comfortable using complex web applications. Instead of building a feature-heavy dashboard, we shifted toward a WhatsApp-first experience while keeping the dashboard for onboarding and more advanced management.

Another challenge was combining multiple variables: market prices, weather, logistics costs, and shelf life, into recommendations that remain easy to understand and actionable.

Accomplishments that we're proud of

Designed a practical solution focused on preventing food loss, not just improving distribution.
Combined multiple agricultural data sources into one decision support platform.
Built a farmer-friendly experience using WhatsApp instead of requiring users to learn a new application.
Focused on helping farmers maximize net profit, not simply selling to the highest bidder.

What we learned

This project taught us that technology alone is not enough.

The biggest challenge isn't collecting data, it's transforming complex information into simple decisions that farmers can trust.

We also learned that accessibility matters just as much as intelligence. A highly accurate AI system has little value if the intended users cannot easily interact with it.

What's next for Tani Pintar

Our next goal is to move beyond rule-based recommendations by training machine learning models using real harvest data collected from farmers.

We also plan to:

Expand support to more crops and regions.
Integrate with additional B2B buyers and cooperatives.
Improve demand prediction and oversupply forecasting.
Build a stronger marketplace connecting farmers directly with institutional buyers.
Measure real-world impact through reduced food loss and increased farmer income.

Ultimately, we envision Tani Pintar becoming the decision intelligence platform that helps farmers across Indonesia make smarter harvest and selling decisions while reducing food waste nationwide.