Show HN: GreenSwapp – API to track CO2 of food products at scale

[ajayvrajan123]

Hi guys! We’re Ajay & Jainit, and we’re super excited to launch GreenSwapp to the HN community!

GreenSwapp is an API that tracks product and recipe-wise carbon footprints at scale, for food products.

Tracking carbon footprints of individual food products is hard because you’d have to trace it all the way back to the farm and account for all the energy inputs till the point of sale. Each such analysis takes about 6 months to do, and in many cases, it is hard to even map this journey. For food products that typically have multiple ingredients (i.e. they are recipes), you have to do this for each ingredient. Supermarkets have about 30k unique products per store. If you do the math, you’ll see that this quickly becomes impossible to do manually, in a reasonable amount of time.

After my Masters’ in Sustainability & Engineering, I (Ajay) worked as a climate & impact consultant for more than a decade. I first discovered this problem 5 years ago (as I ran my own sustainability consultancy), when I saw some of my large food brand & retail clients struggling to identify their highest impact products & suppliers. More recently, we found that food platforms (such as Uber Eats & Yelp) want to track the climate impact of their recipes and restaurants too.

We do the hard work of estimating carbon emissions using a combination of product meta-data / supply chain attributes (such as origin location, farming type, packaging type etc.) and aggregated product carbon footprints from peer-reviewed journals. Our core dataset is a comprehensive list of food types found in supermarkets. We use an ML model (semantic search using sentence transformers) to match queried product names to our standard list of product types and retrieve their median supply chain carbon footprints. This enables us to track impact at scale (thousands of products in minutes), making it easy and cheap for you to track product-level emissions. This would allow you to easily a) report emissions, b) identify high impact products, brands, and suppliers in your inventory, and c) communicate impact with consumers.

What if my product is more sustainable? We identify every supply chain attribute that might reduce a product’s impact (such as organic farming, better packaging material, locally grown) and apply a correction factor (also obtained from peer-reviewed life cycle analyses papers). You can read more on our algorithm here - https://greenswapp.com/api/

You can try our API for free on any milk product (e.g. - Oatly oat drink barista edition) or any type of pizza (e.g. - Pepperoni pizza). Here are the API docs - https://docs.greenswapp.com/. You can get your trial API key here - https://greenswapp.com/api/.

Our full API works with any food product you can find in the supermarket and any recipe. We charge a data license fee of $0.99 per product per year for supermarket products and $0.99 per ingredient per recipe per year for dishes. Here is a cost estimator tool - greenswapp.com/pricing. If you want to look up more than 100 products and 10 recipes, you also pay an annual platform fee (still figuring out the right price point).

We’d love to get your feedback on our API. We are also trying to identify the biggest use case for this product. So please let us know what you’re using it for by commenting on this post - is it to use in a new app you’re building, get your company to use it at their cafeteria, reduce your company’s emissions, offset it, etc.? Any other ideas on how we can make this better would be wonderful too.

Thanks a ton!

[kmanamcheri]

This is great, and looks like there is not many companies which do this so it fills a need. I have some questions :-

1. Can you provide an audit trail describe how you arrived at a given carbon footprint? For example, could you give me a report on which factors of a given product contribute the most to the carbon footprint?

2. Why is carbon footprint THE number to track for environment sustainability? Is a lower carbon footprint ALWAYS the better thing to do? What if a slightly higher carbon footprint allows me to be more sustainable.

3. Is your data license fee for every single query or per product? Does this mean you are data-mining/tracking clients use cases? What is your data privacy policy?

4. If my supermarket is selling Avocados, and I pay $0.99 to get the impact of it, why would I want to re-query this next year. Is the carbon footprint of a given Avocado going to change year over year?

5. Do you have a third-party consultants audit your data? In other words, how do I trust you are not paid for by Oatly milk company to push their products?

[braja]

This is cool. Can you make this work for more complex recipes, e.g., with restaurants? Perhaps import recipes in a standard format to automate? I can see this being a differentiator for catering companies that serve employee lunches (e.g., Forkable).

[ajayvrajan123]

Hey braja14 - indeed. Our API already works with recipes. In fact, you don't even have to enter it in a standard format, or provide us with the ingredients. We tap into a recipe database with 3 million recipes and use machine learning algorithms to automatically find the ingredients of the recipe you want to know the impact of. As you mentioned, we do have a catering company that uses our API. In fact, we're in discussions with a major food delivery app to integrate our API and pilot this in London. They will use it to rate restaurants or recipes as climate-friendly, allowing you to choose better dishes. If you check out the recipe search, do let us know what you think about it. Thanks for your feedback!!

[dirk_poulescom]

This looks really great and perfect timing! I can imagine this will be a must have in the future for supermarkets.

[ajayvrajan123]

Thanks dirk_poulescom.. hopefully the near future :)

[cabbage42]

Super interesting! How do you maintain an up to date database in an industry that is constantly changing?

[ajayvrajan123]

Thanks! Great question - we aggregate carbon footprints from peer-reviewed, published "life cycle assessment" papers on food products. This is a continuous process. Any time a new paper is published on a food product, we find it and update our database with that value. So the actual work of calculating the supply carbon footprint is done by researchers - we are simply data aggregators. We have managed to collect data on fresh as well as processed food types. We have single ingredients as well as products with multiple ingredients. We take into account geographies as well. How? - It took a team of climate and data scientists two years to aggregate life cycle carbon data from every aggregated database as well as journal out there. Do check it out. You can read more on the methodology on our website's FAQs as well. Cheers and thanks for the feedback!

[nikincn]

Hey looks good and it is need of the hour too . Where can I find a list of products available in api ?

[ajayvrajan123]

Hey nikincn - thanks!! We have product "types" available in our database. You can simply query any product name you want - we cover almost all the products you can find at the supermarket. The free, trial API is limited to 1 product type - you can search for any milk product name (such as Oatly Oat Drink Barista Edition) and see the CO2 value, as well as how much better or worse its impact is compared to the average milk.