What it does

AgroAssist aims to enable every farmer to be able to upload pictures of their crops into our webapp and leverage the power of deep-learning to diagnose their crops and connect to the right people for further assistance and explore our products.

We trained a PyTorch neural network based on the ResNet50 backdone which achieved an accuracy of over 93% on the test set. The trained neural network has been deployed into our webapp using ONNX.


The problem we were facing

Connecting to a large user-base is a very challenging task. There was no simple solution to managing and guiding so many potential customers and keeping a track of them.

We wanted to streamline our customer interactions and optimize organic traffic.


Hubspot did all the heavy lifting for us when it came to customer management and product marketing.

  • We used HubSpot Form and ChatFlow to make sure that no user is left in the dark after the preliminary diagnosis of their crops.

  • HubSpot Marketing hub helped us manage our contacts very easily and communicate with our users at scale.

  • The HubSpot Page Builder helped us create professional grade webpages which could be edited and deployed in minutes.


HubSpot made us a lot more ready as a small business to be able to handle a growing customer base reliably and maintain contact with every potential customer on the fly. As a bonus, the HubSpot Page editor helped us build/edit pages in seconds.

Built With

+ 1 more
Share this project: