GreenLens

Inspiration

Consumerism has become much more prevalent through online shopping, encouraging wasteful and unsustainable purchases. We realized that most people don’t choose to make unsustainable choices but are rather uninformed regarding how they can make more sustainable choices and decrease their carbon footprint. It would be so much easier to be sustainable if there was a convenient platform that could inform the consumers of the sustainability of their product choice and also give them more sustainable alternatives.

This is why we came up with GreenLens.

What it does

GreenLens is a convenient and easily accessible Chrome extension that can be used on any webpage. Once the user is on an e-commerce webpage of a product, they can analyze the “sustainability score” of the product. This score is based on specific factors that impact the carbon footprint such as: the emissions from production, the sourcing of the materials, transportation emissions, packaging and the life of the product. Initially, there is a brief summary and explanation for the overall score shown. Then, a breakdown of each of these factors is provided, and when you click on a factor, the user can see exactly why this factor has that specific score for that specific product, and tips on how this score can be improved on future purchases. Additionally, the carbon footprint generated from the production and shipping of that product is also calculated and displayed. We have also provided a way to track the carbon footprint generated by the user by purchasing these products, which is currently calculated by the ‘add to cart’ function but when deploying the extension, it will be based on whether the user actually purchased the product. This function aggregates the carbon footprint and shows the user their overall value. An additional feature we added is the points feature, which allows the user to gain points for more sustainable choices. These points can be used to obtain certain rewards and benefits. This makes the extension more exciting for users and motivates them to continue using it.

How we built it

We used Claude AI to generate the code structure for this extension. Then we ran the extension on Chrome and see if there’s anything we want to add or change. Then, we modified it with Claude or edited the code ourselves on VS Code and ran it again with the extension to see what we should edit next. We continued the cycle until we settled on an extension we are satisfied with. We incorporated the Claude API for the extension to be able to read the info on the product page and assess the sustainability of the product.

Challenges we ran into

One of the major challenges we ran into early on was that one of our team members exceeded the usage limit for Claude AI, which might’ve greatly limited the capabilities of our team. We were, however, able to split up the tasks effectively such that we distributed the work efficiently and were able to successfully complete all the tasks while maximizing the possibilities of our product. Another issue we ran into was with the code regarding the drop-down for the expanded description of factors. The initial couple tries for this feature did not work, and this made it very difficult since we had to generate multiple different prompts until we found the right one that generated a working drop down list with tips and a detailed description.

Accomplishments that we're proud of

We started off with a very rough base model and we were able to come up with a lot of new features that greatly improved the usability and benefits of our extension. Our initial idea was just to build an extension that generates a sustainability score, but we expanded it to provide us with data on the carbon footprint and a detailed explanation of the factors that go into the score calculation. We also added a section for tips and alternatives, which not only allows for monetization after deployment, but also makes it more appealing for customers since they are provided with more sustainable options. Additionally, we added a points feature which makes the extension more game-like allowing users to collect points, which lets us expand this platform to a more social platform, such that users can rank themselves and their friends by points and share their points. The point system also introduces incentives like discounts and coupons.

Overall, this project allowed us to introduce a sustainable shopping platform that informs users about their shopping choices and helps them minimize their carbon footprint and maximize sustainability, overall helping to counter the effects of consumerism on the planet.

What we learned

We learned about the usage of Claude AI, particularly the console and API key and how they can be utilized to build a fully functional product. Specifically, in terms of generating prompts for Claude AI and utilizing the API key in such a way that minimizes cost while still maximizing function. The prompts that we fed Claude had to be specific and clear, so that Claude builds the code with the exact functions we want it to implement. We had to occasionally revise the prompts, making it more detailed and step-by-step. We did also revise the code on our own after it was generated to add certain elements and icons, and to change the overall visuals of the extension. This helped us understand the backend better and learn how to alter the generated code in the way that the extension functioned the way we wanted it to.

What's next for GreenLens

For the purposes of demonstrating the features of the prototype, there were some specific features we were unable to employ fully since we wouldn’t be able to demonstrate it without an actual customer base. The first feature is to distribute the points for the reward system based on data collected from the usage of the extension. We can analyze the score distribution based on collected data of the scores achieved by users on products, and based on that data, we can determine the exact number of points to assign certain scores. For example, certain scores may appear very often whereas other scores are rarely (if ever) achieved. Based on these statistics we can determine the ideal number of points to assign such that it still encourages people to buy higher scoring products, while still making the rewards achievable and incentivizing.

Another factor to consider is that since rewards and benefits are introduced, people would be able to cheat the platform by adding higher scored products to the cart without actually buying them. Therefore, we need to implement a verification process that checks whether the user actually bought the product or not. We couldn’t implement this for the prototype, since if we did, we’d have to buy products to show the working features of the extension. However, if we were launching this product we would ideally implement a checking process that can check for specific content on pages like an order confirmation number. We could also utilize API monitoring tools that detect API calls to payment platforms. Some examples of these tools are chrome.webRequst API which can detect calls to specific domains like stripe, paypal etc. Using this, we don’t need to bypass any security regulations since we don’t need the actual data of the transaction, we’re just monitoring the requests to payment platform domains.

For monetization when scaling the product, we’re thinking of developing partnerships with companies that generally provide products with high sustainability scores. These companies will be recommended under the ‘recommendations and alternatives’ section, as sustainable alternatives to the products that the user is currently viewing. Furthermore, the companies partnering with us would get more credibility regarding their sustainability practices. Additionally, we will set quotas for points so that it acts as a rewards system, and once a specific quota is achieved, the user will get discount codes for the companies that we are partnered with.

The points system can also be expanded into a social platform where users can be ranked based on the number of points they have obtained, and they can share their results with their friends.

Built With

• claudeai
• claudeapi
• css
• html
• javascript
• json