"Food Alliance" project diary


@arthur.tkachenko Read it a second time and it looks like there are quite a few ongoing projects within these thread. Kinda hard to follow up with the posts as I’m a non-techie:)

I used to worked in a supply chain tech firm in China and researched on food safety as well as the integration of technology to improve food supply chain. Do check out http://www.bitsxbites.com/portfolios/ for a handful of food tech startups in China. Bits X Bites is a food tech VC in China.

Anyway what you guys are doing sounds very promising and I personally feel that the online food/ nutritions space is pretty segregated and has huge potential. However, food is incredibly varied hence the difficulty in defining their nutritional content. Compounded by the fact that every individual has different nutritional needs, its a tall order.

1 Like

I think i find another big list of datasets: https://github.com/foodopendata/food-datasets/blob/master/data/food-datasets.csv

@chewweichun94 if you need some help to covert my nedie language and explain something to a non-tech - please tell me what part should be updated - i’ll be happy to explain more.

I don’t hear about Bits X Bites before - thanks for sharing.

I agree that trying to put all things together is hard at food market, but some parts can be updated/automated - this is what we’re trying to do here


@ashley.e.cribb check this: https://github.com/foodopendata/foodrepo-nutrients



yeah here is their API and how it is structured. What would be nice is to have a recipe and then be able to obtain specific nutrients for the algorithm to be able to obtain the score to know the range.

What is going to be important is how to obtain new data after we have a dataset. Then running it through the algorithm and then updating the dataset with the new entry.

I was talking to my mom about the project and she said one thing that would be nice is having a way to substitute ingredients. This way we could make it a better recipe health wise.

1 Like
1 Like

@Natasha looks like I forget to publish this article at some place and it didn’t get a lot of views.
do you think it will be a good article for publishing at hackernoon? -


Right now i’m researching topic related to microservices coding patterns. It’s a great thing, cool movement with a few companies that building tools exactly for developers and a lot of big sharks like Amazon, Microsoft, Google is ready to provide their services for hosting. It also extends the developers workflow, because you can separate not only a code, but also testing/deploy process.

It’s easy to create, not easy to maintain and orcestrate. But the easiest part of doing microservices(aka create and deploy one function that doing only what it should do - like return a recipe) are pretty similar to my modular approach. So it’s like doing a same things as we do right now, but +20% to speed, clean code, less files, easy to code for low-level developers skills.

Just researched another tool that handling event management of communication between user/server and find this simple tutorial: https://www.freecodecamp.org/news/how-to-create-a-realtime-app-using-socket-io-react-node-mongodb-a10c4a1ab676/

it remembers me about restaurants as part of food tech market. It also part of Ashley idea… Quick google search and I find this service provider. It’s complex thing to build from scratch, but very possible to lower the bar for doing this SaaS projects with our modules as well. Right now i’m not ready to do an exact complex project. But partially - maybe later.


Just intersting to read: https://beam.biz/content/113936/EatFun_is_tapping_into_the_massive_homecooked_food_delivery_space/


Yeah. It’s one of the ideas for later for those people who can’t get out to offer the meals. Think of the elderly where it is hard to get out of the house and you have services like meals on wheels to take them food. But what if they are a diabetic?


Danielle’s рroject is pretty similar from the beginning steps that we’re taking with Ashley project.

She also will be able (in my opinion to help us with DB improvements(later) for now(think we will prepare some data for her, using the same scripts, that my team created for my project).
Then we will create a backend system for her project, similar to approach, proposed to both Ashley/Arthur projects.
Then I understand that first frontend pages/views should be also similar logic as well.

What is bothering me – Italian language of data. If we just add it and display - no problems here. But it will add an additional level of complexity if we’ll decide to add localization functionality. It can be solved with small steps, by improving my code modules, but it’s definitely a part that I know how to do, but we never started it. Only in my head for the future. It may not a big problem but will take additional time to do.

There a lot of great work was done related to localization software. one of the latest frameworks came from facebook. But again, it requires some time for research. https://github.com/facebookincubator/fbt

Frontend - that will display data and search/filtration functionality.

Ashley – how we can, need to implement search module at your project - but please keep it simple. Because the whole search is another big topic to discuss together later.

Right now I move only one teammate - she will work with DB+backend system. it will be slow from the beginning, but we can extend team, working on that part in the future.

Danielle --> will you be able to export Italian DB into JSON file or CSV file to me? This will save some time for me and for my teammates.


Ashley -

  1. Ok, your data in the process right now. With my next approach that data will be transferred into the backend, based on graphql server(before uploading data into a database, we will create a scheme that will work with that data should be improved, extended, in order to meet your project requirements)

  1. We can discuss a few blocks from your schema. But not all of them. Because MVP is our goal. We cannot think about all the aspects.

  1. We need to split your functionality into small and understandable blocks/modules. They should work independently. It will help to limit our work and start to work on a few modules in parallel and independently.

3.1 If we will do good documentation like as part of our collaboration - it will save a huge amount of time of my team members.
So let’s start to work together closely and have a lot of small chitchats, that can make a picture of your project as clear as possible.

3.2 Think about what will be next and how to explain it to developers. Next steps, assuming that they don’t know a lot about food data theory. So anything should have a detailed explanation.


This company is an actual example of how food tech moving forward from just a software solution into combination of software + hardware. I think it’s a how things are working right now. A lot of just software companies start to build their own(produced in China)hardware. I think it’s trend right now




Something to keep an eye on


I agree on chefling … though they are talking about smart appliances the major tools for that … front end and back end are correct … we can also look at Whisk https://whisk.com/ … they have built pretty much everything but we have to look at the business model.

One of the items I am thinking is to redo the food algorithm to our needs … so we need to port it to JavaScript or keep it in R and tweak it … but I do like the format of it just that we don’t need all of the items she used for ours … though I recently learned that in a diabetic educators world if you are a man or a woman it makes a difference


Search model.

Elasti search with the database … with auto complete and a mechanism to say whether it’s a recipe or premade food or a restaurant. Usually by a check box etc.

Will be a main function of the UI on the front.


I didn’t implement ElasticSearch before. So I would think that this is a complex functionality for MVP. But if you have time - you can do it on your own - i think our datasets will be enough for start for it. I’ll be happy to have it as you mentioned.



We don’t have to of course … the issue I was trying to address was speed … usually search can take seconds longer than a person using the tool afford and will uninstall it. Just thoughts of course


i like an idea to have a separated search instance with a great index(good speed).

i just didn’t work on it before. so if you want it to have right now - try in yourself and show me what you’ll have. it can be a good part of functioanlty for each project that we’re talking here


Alrighty I’ll make a database and then run the tutorial so we can index it unless we can do that with a GitHub API so we can index some records … also will be starting to work on the front end with figma … please send me a link to the template for invision so I can port it and then work on the screens / functionality


use one of our datasets from https://github.com/GroceriStar/sd/tree/master/generator/projects/USFA

Please do, but frontend is a big topic(huuge). So i think we should discuss details before doing something. So if i can vote - for me elastic search is more important thing than a frontend. especially while i still don’t finish alpha documentation for your project

tell me more about what you want to accomplish with Figma and why you think it’s necessary to create it from scratch. also frontend for what type of “entity” you want to do. Web app, application, etc…

I think you mean this UI toolkit: https://www.invisionapp.com/inside-design/design-resources/scratch-recipe-ui-kit/


Ok when I was talking with a mentor of mine a week ago he suggested to do the front end so that it can help you to know the functionality I was envisioning …

That’s fine I can work on the search functionality. I’ll basically download the database and then put it on AWS and then whip up a elastic search model so that we can start indexing it …

Just was going to do a low fidelity wireframe because the solution we are talking about just isn’t for recipes … so I thought it would help us … so your vote is duly noted … hopefully we can have a kubernetes node so that we can index our food projects

1 Like

Maybe you should show this forum to your mentor? just an idea - i think wecan get a great feedback. I understand the reason of doing frontend.

But if I understand you right - you’ll redo it few times later. From developers perspective - i always tend to have everything document before starting. It’s like a workflow that can save me in bad situations.

ok, curious to see/try your search.

low fidelity wireframe - if you can convert your flow chart that you shared, into

  • then it’s a great idea. but again - i’m sure that it will be changed few times at least. Just based on my experience

kubernetes node is another huuge topic. But "Food Alliance" project diary post it’s a step into this field

nice article: https://sites.google.com/view/sourcesismeathealthy/

Uber is coming: https://angel.co/alpha/stories/uber-will-deliver-food-via-drone-to-san-diego-residents-in-summer-2019-1980