Foodshake April Update | But WHY?

The original update is available on my substack 🙂

See the latest Foodshake changes yourself!

👋 Happy April everyone!

I hope you’ve all been having a great month!

This month’s newsletter is inspired by my brother! He’s a supervisor at a ski hill. He likes to play this game with the students at the mountain — he’ll say to them, “hey little Frankie, everytime your ski instructor says something, ask them but why!?” Then the instructors have to deal with kids asking “but why!?” all day. 🤭

⁉️ But Why? And How?

“But Why?” isn’t limited to just children on the ski hill.

Many people I’ve interviewed about Foodshake have said things to me like:

This looks good, but why is baking powder substituted for eggs?

This is an important question. To fully master vegan cooking and baking, it’s important to know the function of each ingredient in a recipe.

Take eggs, for example. If you know that eggs are leavening, then you’ll have a better idea why baking powder is suggested as a substitute. Foodshake now displays this information.

✍️ Substitution Notes

In order to make Foodshake more educational, I’ve added a section with substitution notes that give further details into why we use certain replacements, and how to use them.

🪅 Multiple substitutes!

The final change for this month was based off of some user feedback:

I think it’d be cool to have possible substitutions. Like if it says “lentils” it could also show other options.

Ask, and you shall receive. I take user feedback very seriously, so let me know if you have any requests!

(For the more technically inclined: I’m lazy, and am giving myself a lot of tech debt 🙂 I loop over my base “recipe.to_vegan()” function N times [N=2 in the below example] and reformat the output).

…and that’s it for April!

Whenever I talk in a meeting at work I always end with “… and that’s it!” So why not continue the trend and end this newsletter the same way?

I’m always looking for more testers of Foodshake. If this is something that interests you, please reply to this email and let me know.

Thanks for reading, and have a great rest of your day!

Finally, here’s a link to Foodshake.

Continue reading “Foodshake April Update | But WHY?”

Foodshake March Update| Goodbye Gratofu!

I sent this Foodshake update out at the end of March, and thought I would cross-post this here (a bit belated)!

See the latest Foodshake changes yourself!

👋 Hi everyone, it’s been a while!

How have you been? I’ve been working away on Foodshake – so much so that February flew by, and didn’t send a newsletter out that month!

Moving forward, I’ll update you once a month on my progress with Foodshake.

So what have I been up to, this past month and a half (ish) ?

🎉 User tests!

TL;DR Foodshake’s conversions have improved!

I’ve had the pleasure of speaking with many of you who have volunteered to try out Foodshake. It breaks easily, and the substitutions sometimes don’t make sense.

Here’s what one tester’s feedback was:

So my notes for this is that I have no clue what gratofu crumbs are

And my response: yes … I have no idea what those are either.

What the heck is gratofu?

This was a *face palm* moment for me, and this is what I realized:

  • The kind person who tested the app had graham crackers in their recipe. GraHAM crackers.
  • Foodshake converts any word it sees that is non-vegan (think eggs, dairy, any meat), to its vegan counterpart.
  • The conversion for ham was tofu: GraHAM —> GraTOFU.

Of course, this is what was going to happen!

(For the more technically inclined, I was doing a simple find and replace because I’m lazy. I have now solved this in a slightly better way using a regex).

Continue reading “Foodshake March Update| Goodbye Gratofu!”

Foodshake app: My motivation

I sent this newsletter out via Substack, and I thought I’d cross post it here 🙂

I hope you’re doing well! Here in Toronto we had a big dump of snow recently. As a result, I haven’t been out as much as I usually am, but it’s given me time to think through more points about Foodshake that I’d like to share with you.

My backyard on our snowiest day so far this year!

The goal of Foodshake is to enable you to convert any non-vegan recipe to vegan. With this in mind, there are a few key components I care deeply about with regards to this app.

I care about creating vegan recipes that:

  1. Are as close to the original recipe as possible. The non-vegan version and the vegan version should taste as similar as possible.
  2. Have accessible ingredients. I want anyone to feel they can make the recipe with ingredients they have heard of before, and that they could find and buy easily without stretching their wallet.
  3. Lean towards whole foods rather than processed foods. There are many products on the market nowadays (Beyond Meat, Gardein, Chao cheese, etc) that allow you to easily swap out non-vegan products for their vegan counterparts. However I care about steering folks towards a whole foods diet, as I believe this is a more sustainable approach to veganism in the long term.

I would love to hear from you – what do you care about ? What is the one thing Foodshake could do that would make you say “wow, my life is so easy now!” ? Comment below, or reply to this and let me know!

Check out foodshake at!

Thoughts on 2021, and the birth of my app: Foodshake!

2021 was a transformative year for many of us, myself included. Covid-19 vaccines arrived as the virus mutated, the supply chain faltered, an NFT was sold for $69 million USD, and at least 4 million Americans participated in the Great Resignation. I couldn’t find statistics on the number of Canadian participants, but I’m one of them.

Being the millennial that I am, I realized I wanted to do something with my work that would truly help people. This is, in part, how Foodshake was born.

What is Foodshake?

Foodshake enables you to convert any non-vegan recipe to vegan.

(Please let me know if you have questions – I’m working on iterating on this description).

Why do we care?

In the absence of Foodshake, how does one convert non-vegan recipes to vegan? I spoke with many vegans and friends of vegans this fall. This quote from one such conversation is a great summary:

I would always look first on the internet. I have so many vegan recipe books now. I still often go on the internet. No matter how many books I have, it’s still faster.

While going on the internet to learn to veganize recipes if faster than reading a book, there’s still room for improvement. I think converting a recipe to vegan should be as easy as: copy, paste, and click.

Who is it for?

Foodshake is for vegans, and the friends and family of vegans. It’s for people like my Dad, people who have good intentions to explore vegan food, but have recipe anxiety. People who have been cooking their family’s spaghetti sauce for years – but how do you make it vegan? Should you use tofu, lentils, or Yves veggie ground round to replace beef in spaghetti sauce? The options feel endless! It’s too much! Foodshake will remove this ambiguity and anxiety and will enable you to quickly convert any non-vegan recipe to vegan.

Product Positioning: April Dunford’s Systematic Approach

Hi! It’s been a while 🙂

Here’s  a little update. I’m currently working on building an app. It’s a recipe app. I’ll leave it at that for now.

To be completely transparent with you, I have no idea what I’m doing. Building an app doesn’t just entail the technical component – if you want to do it right, you need to know: software development, product management, marketing, and other aspects I’m likely missing. Currently I can check one of these boxes off : software development. For the rest, I’m in the dark.

I’m fortunate to have a good friend and mentor to guide me along my journey. He told me in one of our recent meetings that now, after about 3 months of customer interviews, and a working beta, it’s time for me to think about my product positioning.

I have no idea where to start with this (beyond the useful tips from my mentor). I came across a 2019 talk by April Dunford from Tech TO.  Here are my findings.

Components of Product Positioning

April comes from an engineering background, which I very much appreciate. As such, her approach to product positioning is systematic and scientific. According to April, the components of positioning are:

  • alternatives (competitive alternatives)
  • market category (think: am I a CRM? or am I a CRM for investment banks?)
  • unique atributes (features of my product that the alts don’t have)
  • customer segments (what customers actually care about this- who is it for)
  • value (what value customers get, the “so what” for the unique attributes)

And it’s not just a matter of checking each box one by one. You need to do these in the right order. If you start in the wrong spot, you get positioning that isn’t good positioning – it’s not competitively differentiated. April learned that ordering matters through experiments when she taught classes on product marketing to budding startups.

The order April recommends is:

  1. Competitive Alternatives:
    1. competitors, or what people might just do otherwise
    2. go to your best customers and ask them what they would do if you don’t exist
    3. If you start with the wrong comparable, you’ll get to the wrong definition of what your value is
  2. Key Unique Attributes. What do I have that’s unique?
  3. What value do these unique things add? What vlalue?
  4. Who cares a lot about this value, and why?
    1. This is the customer segmentation.
  5. If I’m trying to communicate this value to these people, then what am I?
    1. I.e., what market do I intend to win?

Even knowing that it’s possible to approach product positioning with a systematic process makes me feel somewhat comforted.

Black Lives Matter: resources for machine learning practitioners, researchers, and allies

Black Lives Matter! As the diversity and inclusion lead at Aggregate Intellect, I felt that it’s important to make our stance clear. At Aggregate Intellect, we recognize that this is not an issue for the Black community to solve alone. This is an issue for all of us to solve together. As a community of machine learning practitioners, we are all equipped with a unique skill set that allows us to make a difference. This post outlines ways you can contribute as a technologist, and how you can contribute as a general member of society as well. Special thanks to Suhas Pai, our NLP Stream Owner and the director of our weekly newsletter for putting together a large portion of these resources!

Continue reading “Black Lives Matter: resources for machine learning practitioners, researchers, and allies”

Some thoughts on comparable companies analysis and machine learning

I’ve been reading up on investment banking lately for work, also out of interest. The first chapter in Investment Banking: Valuation, Leveraged Buyouts, and Mergers and Acquisitions by Joshua Pearl and Joshua Rosenbaum talks about comparable companies analysis, also called trading comps. It’s one way that investment bankers go about valuing private companies that are about to go public, or those that may merge or be acquired by another company, for example.

Automating comparable companies analysis

The process seems fine, it involves data collection, then essentially comparing the “target” (the company you’d like to value) to those in a “comparables universe.” It’s all manual work, with the most difficult part being amassing the universe of comparable companies. The rest of it essentially involves simple addition, subtraction, multiplication, and division. And it makes me think… why can’t we just automate it with machine learning?

Continue reading “Some thoughts on comparable companies analysis and machine learning”

Can algorithms be patented? A look at Amazon’s item-item collaborative filtering system

In 2003, Amazon revealed their secret to the world: their item-item based collaborative filtering algorithm! In their paper, Recommendations Item-to-Item Collaborative Filtering, which has ~6000 citations on Google Scholar as of October 2019, they share the details on their recommender system that drove their website’s revenue for years. 

There are 3 interesting problems that are solved in this paper. Those are:

  1. how to represent items
  2. how to compute similarity of items, and
  3. how to scale the solution.

And, the most interesting part of this is: they patented it.

Before we discuss the patent, and ask some questions, let’s step through each sub-problem and how they solved it.

Figure 6 in Amazon’s item-item collaborative filtering patent. This figure describes their recommendation product: an email. Quite innovative at the time!

Continue reading “Can algorithms be patented? A look at Amazon’s item-item collaborative filtering system”

Order Matters: Alibaba’s Transformer-based Recommender System

I wrote a blog post describing Alibaba’s new recommender system that leverages the popular Transformer architecture. It’s a great example of the intersection of NLP and Recommender Systems. I originally posted it on the AI Socratic Circles blog, which I manage, so please feel free to navigate there if you’d like a full-page experience, and if you’d like to check out some of the other content. For those who prefer to stay on this page, I’ve embedded the original article below. And thanks to the two editors of the post, Susan Shu and Omar Nada.

Enjoy the post!

A three-pronged approach: how to get a job as a data scientist


Before I became a data scientist, I spent a lot of time Googling “How to get a job as a data scientist” and browsing r/DataScience. Maybe you’re trying to become a data scientist, too, and you’ve somehow landed on my blog. Welcome! 

I’m writing this article for 2 reasons:

  1. I want to help you! I was lucky to receive a lot of help from generous members of the data science community when I was first starting out. This article is one way I can give back to people who are in the position I was in. We all have to start somewhere! 
  2. I find I often give the same advice over and over again to people who reach out to me to chat, so I’d like to put my thoughts down in one place. Hopefully this article can be helpful to you in some way. 

Feel free to reach out to me if you have any further questions and if I can help in any way!

Continue reading “A three-pronged approach: how to get a job as a data scientist”