Groceroo

Save money on groceries with real-time price comparisons across supermarkets. Get the best deals, a smart shopping list, and AI-powered recipes. Shop smarter, save more—effortlessly!

Comment

Inspiration

In this economy, grocery shopping feels like a game of survival. One week, your favorite cereal is $5, the next, it’s practically a luxury item. Wouldn’t it be nice to know exactly where to get the best deal without hopping between apps or running around town? That’s where Groceroo comes in.

What it does

A Smart Grocery List

First and foremost its a grocery list which lets you select items and the web app will show you the cheapest options available from each grocery store.

It will then generate separate grocery lists for each supermarket so you can save money. Don't want to go on two trips? No worries, it will generate just the one for you where you can save the most.

AI Powered Recipe Maker (to be implemented)

Ever craved to make a certain dish? Don't know how to cook that dish or where to get the ingredients? Groceroo, with its AI powered recipe maker, will generate a recipe for you and add the ingredients into your grocery list.

How we built it

The web app is built on Python Flask for its backend and React with Node.js for it's frontend. Using publicly available APIs from grocery stores such as Coles and Woolworths, it will grab the latest price and availability from each.

How to run the project

  1. Make sure you have installed Langflow locally. Follow the installation guide.
  2. Run Langflow and import the Langflow Local.json file.
  3. Open the project directory.

  4. Install the required dependencies: pip install -r requirements.txt

  5. Start the backend: python app.py

  6. Navigate to the frontend directory: cd front-end

  7. Install and start the frontend:

    npm install
npm start

Challenges we ran into

Both Astra DB and Langflow are entirely new technologies for us. Therefore, getting both to work for the first time and integrating them with our web app was quite challenging.

The first roadblock we faced was figuring out how to make it possible for us to search for similar products from each grocery store. However, we figured out a workaround: Instead of storing every product in a single collection, we divided the product data we gathered based on the source grocery store and stored them into separate Astra DB collections. Then, in Langflow, we created a flow to retrieve the products from those collections simultaneously and use them as part of our RAG workflow.

Accomplishments that we're proud of

  • We built a fully functional app from scratch, despite zero prior experience.
  • Successfully tackled database and AI challenges, making the app smarter with each iteration.
  • Even though some features aren’t fully implemented yet, it’s amazing that we got this far, with no prior front-end or design experience (in less than 48 hours).

What we learned

Building an app is like grocery shopping on a budget—you think you know what you need, but halfway through, you realize you're missing key ingredients. We started simple but quickly realized AI, databases, and real-time pricing were trickier than expected.

Teamwork is everything. No one knows it all, but when you combine different strengths, the impossible becomes doable. Some of us were great at backend, others at frontend, and some just had a sixth sense for debugging.

Tech will throw curveballs. Grocery stores don’t name products the same way, APIs can be unreliable, and AI isn’t magic, it only works as well as the data behind it. We had to constantly adapt and rethink our approach.

Agile mindset > Perfectionism. Instead of waiting for the perfect version, we built, tested, broke things, and improved fast. Turns out, progress beats perfection every time.

The best way to learn? Dive in. We had zero experience with Astra DB and Langflow. But by struggling, debugging, and iterating, we figured it out. Nothing beats hands-on learning.

What's next for Groceroo

Implementing further AI features. Aside from only tracking the price, the app can also track calories and essentially be a calorie meal planner. The AI would recommend protein rich and low calorie recipes to achieve your goals.

Another feature to implement within the app is to include more groceries, not only the major ones but possibly crowd-sourced prices from smaller supermarkets.

A future plan is to also generate a journey trip to see how long and how much it would cost you to go from one supermarket to another. To see if its even worth the trip.

Built With

+ 5 more
Share this project:

Updates