I have worked as Head of Technology for an Agritech Ecommerce startup since 2019. After completing my engineering, I worked as a Software Developer for various organizations. I could see a broader aspect of the AIML for different industries, and so many organizations were working in this field to make decisions based upon some facts and logic. With an experience of around 11 years, I have prior knowledge of mobile application development. Before this, I was working in the organization named FarmX.
Since many organisations are working in the sector with the help of existing data and making decisions, I have thought of working in this area, so we can create some AIML projects to help our customer and internal team make decisions. Some of the projects could be like recognizing products from images, finding out the quality of products from images, looking into the trend of pricing of a product, etc.
The problem we were facing is that we could not predict the product’s prices after 3 to 4 days, so if the price falls, we have to sell our agriculture product at a loss. This was severely affecting the margin. Suppose we have to send a truck of onion from Nasik to Delhi, which takes 3 days. Then we are not sure that after 3 days, the onion price will fall or rise. So we need a solution to predict approximate product price from past data.
Now when we know what could be the future rates and if it is going to be a profit then only we sell the product. I have used the AIML technique and ExtraTreeRegressor to predict the Onion and Potato prices for the Bangalore and Delhi market. Regression and Supervised learning techniques helped me out in this. I have used various Supervised Techniques and models to find out, but no one was giving perfect results. I searched for another model named Extra Tree Regressor, which improved the results when I implemented GridSearchCV. Improved the results to reach an accuracy of 96%.
I recommended and tried various solutions that used Neural networks, NLP, and supervised learning techniques but practicing different solutions, I came to know that demand and supply matter a lot, and cracking that thing using regression techniques helped organizations a lot.
My organization has a great significant impact on this solution. Now, losses have been reduced to a great extent to get better margins with the help of finding the range of prices in upcoming days. This exercise gave me immense experience and boosted my skills and confidence regarding data analysis and AIML. I have started with the next project of finding plant disease from its leaf and suggesting suitable pesticides for the same. I was unaware of how the AIML works, and for this project, I needed excellent in-depth knowledge of AIML. During this project, I referred to lectures many times, and those helped me to accomplish this project, so helping my organization.
