United in college and knotted by passion, this is the story of two young engineering students, Tanish Karwekar & Sankalp Pandey, willing to help farmers across Madhya Pradesh with their AI tool, KrishiGPT.
Born and brought up in a state like Madhya Pradesh, both (Tanish and Sankalp) are empathetic towards problems faced by farmers. Realising this, they decided to do something within their capacity and innovated something that can really help farmers across India.
Here is another story of young innovators proving India's Got Intelligence.
Introduction (Personal / Team story)
Tell us something about your story. What's something fascinating about you?
Tanish - Have you ever had a dream that was so vivid, so real, that you just had to make it happen? For me, it was a literal dream. One night, I dreamt of owning a restaurant chain called 'The Flying Chariot'. It just stuck with me! I was on a path to join the army, but when that didn't work out, I found myself in an engineering college, studying Electrical and Electronics. And that's where I met Sankalp, my classmate and partner-in-crime.
My story (Sankalp here!) is a bit different. I come from Ujjain, and honestly, no one in my family had a background in engineering or even a degree. I come from a very humble background. My father worked as a labourer. My inspiration? Movies! Seeing characters in films like Tony Stark and Mr. Robot hacking and building cool stuff pushed me towards engineering. I started learning about ethical hacking and cybersecurity at a local institute in 10th grade, bunking school to learn computer hardware and networking.
- But then, lockdown happened. My learning stopped, and I pivoted to video editing to earn some money freelancing. I got pretty good at it, helping a YouTuber grow from 50,000 to 3.6 million subscribers! But my mentor snapped me out of it. He said, "Bhai, you came to learn engineering and coding, what are you doing in video editing?". He was right. I was just there for the money. So, I switched my focus back to engineering.
I didn’t have a laptop, so I taught myself C++ on my phone. With the money I earned from video editing, I bought a second-hand laptop and dived into coding.The funny thing is, Tanish and I started exploring AI individually. We didn't even know the other person was into it!
Then, one day, we connected, and it was like magic. We just clicked. Now, we're so deep into our projects that we practically live together. I pay rent for one place but stay and work with him, often pulling all-nighters, working for two days straight without sleep! Together, we decided to channel our energy into something that could make a real difference, and that's how KrishiGPT was born.
Problem Overview
What is the problem you are trying to solve, or what innovation are you working on?
We both come from families with farming backgrounds, so we’ve seen the struggles of farmers up close. Imagine a farmer in a small village of just 50 houses, miles away from the next town. His crops are dying from a disease he can't identify, and there's no expert around to help. This lack of technology and expert guidance is a massive problem in rural India. Farmers are just now getting used to smartphones, so any solution had to be incredibly simple and easy to use.
We wanted to tackle several key problems with one application:
• Crop Disease: How can a farmer quickly find out what’s wrong with their crop and how to fix it?
• Market Access: How can a small-scale farmer learn about export opportunities and grow their business?
• Fair Prices: Farmers often sell their produce at low prices because they don’t know the current market rates.
How could we help them get a fair price for their hard work?
• Information Overload: Information on government schemes, subsidies, and the best fertilizers is scattered. We wanted to bring it all into one place.
Who will benefit from your work?
This will benefit farmers at indivdual level. They can use KrishiGPT in their mobile phones.
Motivation
What motivated you to work on this problem or innovation? Why are you so concerned about it? Share your story.
Our motivation is deeply personal. I (Tanish) have seen many family members struggle in farming. And I (Sankalp) saw my own uncle’s entire crop get destroyed by yellow fungus.
The helplessness you feel is immense. When we developed a prototype of KrishiGPT, I gave it to my uncle. I didn’t even have to explain the interface; he figured it out himself!
He used the 'Ask by Voice' feature, got his answer, and his face just lit up with relief.Seeing that joy, knowing we had built something that could genuinely help, was the biggest motivation for us. We went around the village, gave the app to more people, and their positive feedback was overwhelming. That feeling, when you create something that solves a real problem for someone you care about… it's priceless. We knew we were on the right path.
Solution Overview - The AI Use Case
Our solution is KrishiGPT, which we like to call a multi-language, voice-first, and image-powered AI assistant for farmers. Think about it:
• Multi-language: We have implemented over seven languages so farmers from every corner of India can use it.
• Voice-First: Many farmers may not be comfortable typing. So, they can just ask their questions out loud.
• Image-Powered: A farmer can simply click a photo of a diseased plant, upload it, and the AI will diagnose the problem and suggest a solution. It can even scan a handwritten expense bill and digitise it to help manage finances.
• Advanced Features: We've also added new features like
- Hyperspectral Analysis
- Predictive Pest Alerts
- Precision Input Advisor
- Crop Health Dashboard
- Live field health maps > NDVI/NDWI trends > Anomaly detection alerts
- Smart Irrigation Planner (Soil moisture + weather forecast > AI creates optimal irrigation schedule) &
- Yield Optimizer
The AI also gives recommendations, like the best time to buy fertilizers to save money. It even predicts market price fluctuations using historical data, like stock market indicators, so farmers know the best time to sell. And the coolest part we're working on? Using Augmented Reality (AR) to let a farmer see a 3D model of a crop on their empty land, helping them visualise their future harvest!.
Technical Details - How AI is Actually Utilised
So, how does this all work? Let me break it down. We started by using Firebase, which helped us with hosting and backend services.
For the AI part, we utilised several of Google’s APIs, including Gemini Pro and Vertex AI Vision, which are well-suited for processing images, text, and voice inputs.However, we didn’t want to rely solely on these APIs, as they can become expensive.
So, we started building our own brain for the app. The core of our "Crop Doctor" feature is a Convolutional Neural Network (CNN), a type of Machine Learning model. We trained this model on a massive dataset of over 500,000 images of various crop diseases, which we sourced from government portals like Vikaspedia.
For predicting market prices, we used other ML models like Linear Regression and Support Vector Machines, which are excellent at analysing large historical datasets to find patterns. Essentially, we are building an application that can eventually learn and grow on its own, just like a child.
Challenges & Failures
Our biggest challenge has been dependency. We were relying heavily on Google's APIs, but the free credits ran out, and our app is currently down. This pushed us to build our own ML models, but it’s like we, two electrical engineering students, are trying to create something that can compete with what top Google engineers have built! It’s a huge task, but a necessary one to make our project self-sustaining. Another challenge is funding.
We have a vision, but we lack the resources to scale it up, get a trademark, and deploy it on the Play Store. We are constantly looking for investors who believe in our mission. We also learned the importance of simplicity. During a presentation, we realised our app was slow on desktops because we had designed it entirely for smartphones, which is what our target users have. Making things simple and user-focused isn't just a feature; it's an innovation in itself.
AI Readiness Skills
How good are your Collaboration skills? Primarily, it’s just the two of us coding and building this together. But recently, we found a great mentor, Mr. Manoj Mali, who is a team lead at ESDS Software Solution in Nashik. He saw our project at a hackathon and is now guiding us to take it to an industry level.
How good are your Critical Thinking & Creativity ? We believe our solution is currently one of the best out there because of its extreme simplicity and user-friendly design. Even our mentor, who was working on a similar idea, was impressed by how intuitive our UI/UX was compared to his more traditional, "orthodox" version. We both come from a design background—I (Tanish) am the head of magazine design in our college—so we always think about the user experience first.
How good are your Online Learning skills? As electrical engineering students, everything we did for this project was self-taught. We are huge believers in learning by doing. I (Sankalp) learned extensively from YouTube channels like freeCodeCamp, which offers free, high-quality courses taught by experts from Harvard and Google. I (Tanish) prefer learning by reading research papers from authentic sources like ResearchGate and IEEE, which helps me understand complex topics in depth.
How good are your Online Reputation skills? We know how important it is to share what you're building. We’ve shared our journey and the initial development stages of KrishiGPT on LinkedIn, which helped us connect with people and get feedback.
Impact So Far & Vision for the Future
The biggest impact so far has been on a personal level—seeing the smiles on the faces of farmers in our village who used the app. Our vision is to make KrishiGPT available to every farmer in India. With support for multiple languages and a simple interface, we believe we can easily get over 10,000 downloads once it’s on the Play Store. We want to create a full ecosystem by involving traders and fertilizer companies to make it even more useful. This isn't just an app; we believe it can be a revolution for Indian agriculture.
Your Advice to Fellow Innovators
Tanish: My advice is to focus on simplicity. In the age of AI, don't just build complex things because you can. True innovation lies in taking a complex problem and making the solution incredibly simple and accessible for everyone. And remember, how well you communicate your idea is just as important as the idea itself.
Sankalp: Don't get trapped into thinking you need expensive courses to learn AI. There are amazing free resources like freeCodeCamp out there. Also, surround yourself with the right people. Having a partner like Tanishk, who is on the same wavelength, makes all the difference. Your community matters a lot.
Mentor, Support & Inspiration
Besides our new mentor, Mr. Manoj Mali, we draw inspiration from different places. For me (Sankalp), it’s a bit childish, but my hero has always been Tony Stark. I started my coding journey with the dream of building my own JARVIS. And just the other night, I actually built a basic version in half an hour—something that would have taken months just a few years ago!. I’m also inspired by Leonardo da Vinci, a master of multiple fields.
For me (Tanish), my inspiration is Steve Jobs. His obsession with creativity, design, and simplicity is something I deeply admire. We are both driven by these figures who dared to think differently and change the world. And that's exactly what we hope to do with our work.
Our Project
Anyone can assess to our project here :- KrishiGPT
So this was the story of young brains who are willing to solve problems by doing. Team of India's Got Intelligence wish them all the best for their future endeavours. Good luck! Keep proving India's Got Intelligence.
