ATOMS: smart inventory management tool

Inspiration

The covid pandemic has had an unprecedented impact on healthcare workers. It is essential now more than ever to help frontline health professionals free up time to focus more on patient care by optimizing the medical supply chain using whatever technology can support.

What it does

Automate the process of managing inventory. Predict and recommend future requirment to ensure there is no deficit of any medical supplies or drugs. Intutive app to order/keep track of current orders/shipments. This will also help with bulk orders and fleet management to save time, energy and money for all parties involved.

How we built it

ATOMS is a full stack web application powered with ReactJS for the front-end, and Flask, SQLAlchemy, PostgreSQL powering the back-end infrastructure of the application. Our Machine Learning model is trained using dataset of medical supply/drug store(russman) data from Kaggle to predict retail forecasting for application in fields of bulk order that can help in efficient fleet management.

Challenges we ran into

1. Understanding medical supply usecases and challenges faced in medical supply chain.
2. Gathering data to train machine learning model for retail forecasting.
3. Optimizing our ML algorithm to train our model with the limited data available.
4. Learning and working with new team members, on new technologies.

Accomplishments that we're proud of

1. Mission first approach. We are happy with exploring the impact technology can have on helping and supporting medical professionals with automating routine tasks.
2. We're happy with our approach to the hackathon. We identified the real life problems that could be addressed and analyzed the datasets to brainstorm identify areas that needed improvement and how we can approach to help solve the problem.

What we learned

Our team had a unique experience attending this hackathon(first for all of us), meeting new people, discussing different ideas and learning new technologies. As part of this project, our team members have gained knowledge of full stack web development where we explored ReactJS and NextJS for front-end, Flask for backend api's, PostgreSQL for managing the database, research on medical datasets, cleaning, identifying relevant information on top of the dataset and finding optimal algorithms for prediction. Also, we've learned and experienced how fun a hackathon and working on a fun project as a team is! :)

What's next for ATOMS: a smart inventory management tool

1. Enhancement of Machine Learning model. This includes optimizing the algorithm further and also finding better datasets to train the model. In a medical retial quantity perspective each product has unique criteria such as expiry, cost, availabliity, demand, demographic, season etc so that should be kept in mind when working on our model.
2. We can work on how this format can be widely used to improve planning in medical supply chain thereby helping suppliers to optimize their inventory maintenance based on expected demand and improve management and deployment of their fleet to meet demand while helping climate by saving energy and time.
3. Enhance our web application to have a more complete MVP that can be open-sourced. So far, we have a proof of work on how our application can help with the inventory management.

Built With

• flask
• machine-learning
• postgresql
• python
• react
• sqlalchemy