The design of recipes on the internet

This is the story of how I started Recipe.Site and what I built over the first three years. Recipe.Site is the first software company I’ve started (though I’ve been designing and building software for years). I haven’t raised any money and currently own 100% of the company. Half of the time I don’t know what I’m doing, but I can say, without question, I’m enjoying working on this more than anything else I’ve worked on.

What is Recipe.Site?

Recipe.Site is a recipe-specific CMS and publishing tool for recipe creators. To understand what that means, look at these three websites:

• https://hwoolee.com/
• https://www.cheffinwithzach.com/
• https://cookbooks.katieparla.com

They were built with Recipe.Site. The first two are recipe subscription websites — H Woo and Zach publish recipes weekly (some are paywalled, others are free). They also publish non-recipe posts, send newsletters, respond to comments, create recipe collections, add items to their “Pantry”, all through Recipe.Site. Anyone can subscribe monthly or yearly to access the paywalled content.

The third site is the digital version of Katie Parla’s most recent cookbook, Rome. It’s what digital cookbooks should be — properly formatted, interactive experiences with comments, scaling, conversions, videos, bookmarks, filtering, etc.

The most important way these three sites differ from other recipe websites is the recipe experience. For Creators editing and publishing recipes, and for people cooking them, the recipe experience is more functional, refined, and (dare I say) beautiful than anything else out there. To illustrate this, let’s take a look at some other independent sites (not platforms like New York Times Cooking or Mob):

Alison Roman (Simple Fruit Tart recipe)
This site was built on Ghost by a design and development agency. It cost $100–200K. A lot of love clearly went into the design, but the recipe pages don’t offer scaling or measurement conversions. Your screen falls asleep while you’re cooking. There aren’t proper user accounts (the bookmarking feature “is only visible on this browser. Clearing your browser or switching devices will remove saves.”).

Molly Baz (Pickley Pulled Pork recipe)
This site was built on WordPress (with a Memberful plugin for subscriptions and newsletters), by a design and development agency. It also cost $100–200K. Molly pioneered the recipe subscription club, the branding is beautiful, but the recipe pages have the same shortcomings as Alison Roman’s.

Caroline Chambers (Apple Chicken Farro Salad recipe)
This is a Substack newsletter/site. Same deal, arguably worse: unless she were to use the recipe card feature (most people don’t), the recipes have no formatting at all, and even then they don’t have any utilities.

Cabagges (Chicken Soup Rice recipe)
This site was built on SquareSpace. I believe they designed it themselves, with development help from a friend. It’s beautiful, and loved by many creators and cooks (I love it, too). But the recipe pages have the same shortcomings.

The Woks of Life (Sesame Chicken recipe)
This is a traditional WordPress recipe site with WPRM for recipe formatting and Raptive ads. Interestingly, these sites have useful recipe features; it’s the ads that make them unusable (Dan Pelosi’s site is a great example of how nice a WordPress recipe site can be without ads). But ads or no ads, they fall short on the Creator experience — using WordPress sucks.

So that’s the landscape, and why I started Recipe.Site. If you were a recipe creator and wanted a beautiful, functional, state-of-the-art recipe subscription site or digital cookbook, and you didn’t have tons of money or weren’t a designer/developer, there were no good options. You could compromise on functionality and use Ghost, Substack, or SquareSpace, or compromise your own experience and use WordPress (and even then, without spending at least $30K, you wouldn’t get a site that looks very good).

Year 1 (November 2022–November 2023)

First prototype

In November 2022, as I was wrapping up the Wikipedia desktop redesign, I started playing around with a recipe CMS + website idea I’d had for a while (more personal backstory at the bottom of this article). I had been making Eggplant Parmesan (using Claire Saffitz’s recipe), and I thought trying to reformat it to be easier to follow would make a good proof-of-concept.

Here’s the prototype I made: Eggplant parm prototype.

Some things I was thinking about:

• Instructions outline — Eggplant Parmesan is an involved process, so I wanted an outline. Something to help me understand the general structure and rhythm — you’re going to start the sauce → do a bunch of chopping and drying → take a bit of a break → very messy frying process, etc. I could form an outline in my head by reading the recipe, and some recipes use section titles to effectively orient the reader. Not a novel concept, just something I wanted to explore.
• Ingredient formatting — I’ve always found it odd that ingredients list their quantities first:

2 cloves of garlic

• When scanning an ingredient list I care most about the ingredient name, not the quantity. Some websites and cookbooks bold the ingredient name to make it stand out, which helps:

2 cloves of garlic

• I thought it would also make sense to put the name before the quantity, so the list is easy to scan:

Garlic (2 cloves)

Olive oil (1 ¼ cups)

• I prefer this format, but I assumed recipe writers would want to decide for themselves, so there’d need to be a setting for ingredient formatting/syntax.
• Recipe summary — I also wanted a clearer summary of the time and effort required. Recipes often do this with timing, difficulty level, etc. I thought a combination of photos (each showing a key step), the number of ingredients, timing, and my outline/map thing, would summarize it well.
• Optional steps — Since making the sauce from scratch is optional, I wanted a toggle to switch between making sauce and buy sauce. I also imagined other variations you could toggle on/off (e.g. “healthy” or “vegan”).
• Timers — There are lots of timed steps, so I wanted inline timers (rather than switching to another app)
• Losing my place — I would often read a step, do it (or part of it), then turn back to the recipe and struggle to find my place in the paragraph. I wanted shorter instruction paragraphs and collapsible sections, so I could close finished ones and keep track of where I was. I imagined this was partly related to experience level; perhaps very experienced cooks just read and remember the whole paragraph.
• Technique videos — Though I didn’t need it for this recipe, I thought it would be cool if steps involving specific techniques (e.g. crushing garlic, chopping an onion) could include YouTube videos of them. That way creators wouldn’t have to re-make videos of generic techniques, and I wouldn’t have to switch to YouTube mid-recipe.

The prototype’s main shortcoming was that I only considered larger screens (I use a laptop in the kitchen) — I hadn’t yet realized that almost everyone cooks from their phones. I tinkered with it for a few weeks, showed it to a few people, and started looking more closely at other recipe sites. I made a few diagrams in Figma to articulate some common pain points, but most of my thinking stayed in my head.

Second prototype

Here’s the second prototype: Saffron Pasta prototype. I was mainly thinking about:

• Mobile — by now I knew most people cook from their phones, so I treated mobile as the default layout and desktop as secondary. I also considered how many people arrive from Instagram or TikTok, and tried to match familiar patterns (e.g. image/video carousels).
• Ingredient quantity lookup/interruption — People told me they often scroll back up to the ingredient list while cooking to check quantities (they see something like “add soy sauce and white pepper” but don’t remember the amounts). My solution was to put the ingredients list in a bottom sheet, so you didn’t lose your place. This one is interesting because it represents an overarching challenge: anytime someone has to leave the recipe page (or go to a different part of it) to look something up or do something, it’s a failing of the interface, and an opportunity to improve it.
• Equipment — Bon Appétit showed photos of special equipment needed for certain recipes, which I liked, so I copied it.
• Experience level — how does a beginner read and cook a recipe differently from an expert? Does a beginner want short steps and lots of images/videos? Substitutes suggested to them? More explanation of what certain ingredients and techniques do?

Maggie

I wanted to talk a professional recipe blogger. I had plenty of experience cooking from recipe blogs, but knew nothing about the creator-side. I emailed maybe 10 people and connected with Maggie of Omnivore’s Cookbook. I paid her to share her workflow and explain her business. I learned that:

• She had to constantly pay for maintenance on her WordPress site (which she didn’t particularly like to begin with, though it looks great)
• She (and everybody else) used the WordPress Recipe Maker plugin for recipe formatting, plus a bunch of others (SEO, collections, etc).
• She was making a lot of money from ads through Raptive, but it took years to build
• She spent most of her time on recipe development and content production, with help from a small team

I considered building a better WordPress recipe plugin. Almost all recipe bloggers seemed to use WordPress (some were on Substack too, but I didn’t know that yet), and changing plugins is way easier than changing platforms. But no matter how good a plugin is, WordPress is still a pain in the ass to use, so that felt like too narrow a solution. I also felt strongly that programmatic ads ruin the cooking experience, so I wasn’t interested in that ecosystem.

…and Jim

The other person I had in-depth conversations with was Jim Liu, a friend and former co-worker from AltSchool. Jim has a special mix ofanalytical thinking, practicality, familiarity with software products and businesses, cultural intuition, and supportive energy. He helped me articulate product approaches and business models, identify my strengths, and stay encouraged. (And he later connected me to Dana, who connected me to Alice, a recipe creator who became my first partner/beta user).

Quitting my job

In February 2023, the Wikipedia redesign launched, I shaved my head, bleached it blonde, and quit my job. I didn’t have a business model or a spreadsheet of numbers, but I had a clear sense of what I wanted to build and a strong intuition that it would be useful to creators. I moved from Vermont to NYC (which I find more energizing) and rented a desk at a co-working place near my new apartment. I switched to an earlier sleep schedule (10pm–6am), and stuck to it closely, knowing that any structure I wanted I would have to create for myself.

I started feeling my way forward — relying on intuition and relationships rather than an overarching plan.

Third prototype

Here’s the third prototype: Lavash Triangles prototype. Some notes:

• I started thinking about the entire site (not just the recipe pages) — navigation, how to represent recipes as cards in a grid or rows in a list, etc.
• I also decided, in addition to the bottom sheet, ingredient quantities should appear inline, within the instructions, e.g.:

add the soy sauce (1 tsp) and white pepper (2 tsp)

Since I wanted scaling, these references had to be dynamic rather than static text, so they would update when the recipe was scaled, e.g. if doubled, the instruction would read:

add the soy sauce (2 tsp) and white pepper (4 tsp)

Ingredient database and structured data

Though not part of the prototypes yet, much of the UX I envisioned assumed an ingredient database and recipes stored as structured data. In short, structured data enables scaling, US/metric conversion, dynamic ingredient substitutions, detailed recipe filtering, and more. I recognized that ingredients were the atomic units of recipes, and that splitting them into component parts (name, quantity, unit, preparation, etc.) and normalizing them against a database would be important. I was interested in the tension between the idiosyncratic way people write recipes and the structure the CMS would “want.” The challenge would be to create a recipe editor that let you write fluidly while still capturing the structure needed for the rich, interactive experiences readers/cooks wanted.

Year 2 (November 2023–November 2024)

Alice

I felt the next step was to find one or two recipe creators without websites and offer to build them one. Jim introduced me to Dana, who kindly spread the word about my offer with a few friends, which led me to Alice. Sidenote: at the time I wasn’t surprised Alice was open to it (who wouldn’t want a nice, free website?), but over the past three years I’ve been surprised by how few people are willing to try something new, which has recast Alice as uniquely open-minded to me (thank you again, Alice!).

Alice had a website on SquareSpace that she wasn’t attached to, and was excited by my prototypes of more functional recipe pages. We also just connected; she understood why I was doing this and wanted to participate. We chatted a bunch, aligned on the broad strokes of the site’s content and look, and I started chipping away at it.

The biggest question was whether to build the site statically and add her recipes by hand, or build a more formal product with a CMS/backend she could use herself. Enter Conor.

Conor, and a name

I was working out of Index in Chinatown. There were a bunch of like-minded, “creative technologists” there, and being around them made me wonder about a collaborator. I was also getting lonely. Starting something from scratch, in an industry I was unfamiliar with, requiring a bunch of skills I didn’t quite have, by myself, without clear direction, was a challenge. Some days felt long, progress seemed imperceptible, and I often doubted myself. My world felt like it was narrowing. I did a short stint as a medical scribe, which added helpful variety and grounding, but was unsustainable alongside the business.

Isabel was one of the first people I became friends with at Index. After showing her what I was working on she suggested I chat with Conor (who was in Italy at the time, on vacation with a friend, which created some nice anticipation). As luck would have it, Conor was a one-in-a-thousand engineer who could do everything from backend to frontend…to UX to graphic design to video editing to woodworking to sewing to dressing well to analyzing art to playing music to making agedashi tofu, and the list goes on and on. And he does all these things at an extremely high level, with supernatural attention to detail, a deep love of craft, and supreme confidence. To say the least, he’s a rare individual, and I was enamored with him. He also seemed popular in the Index community, which was attractive.

We had a lot in common, he was excited by the idea, and impressed by my commitment. Talking with him about the project made me feel excited, buoyant, and grounded. I naively imagined that if he got involved, everything would just start to happen.

I didn’t have a name yet, and in one of our first conversations Conor told me that finding a name is a magical moment, marking a project’s formal existence and setting things in motion (he put a lot of emphasis on names — later on he said, half-jokingly, that he could sell anything with the right name). I started a list of names. I liked the idea of a general term for a container or collection (e.g. Arena, Index), and I also liked the idea of an extremely literal name (something with recipe in it). One quiet Saturday at the office, I was adding to my list of names and checking domains when I stumbled upon “recipe.site” (currently pronounced “recipe site”). I liked the simplicity and specificity, and that the name and domain could be the same without sounding like a domain name. I liked that it felt quiet and generic, because I’d started thinking a defining feature of the product might be that it’s behind the scenes — a tool, scaffolding for your content — rather than a quasi-media company like Substack.

I remember my excitement, a few days later, telling Conor: I found a name.

In December 2023 we started a trial collaboration: he did some tightly scoped engineering work, which gave us a chance to try working together and think about what a more formal arrangement might look like. A few months later we agreed he’d get some equity. We were having fun. Staying up late, working on the weekends, eating Chipolte, building something we both loved. However, looking back, there were early signs that our working styles weren’t compatible. Plus, I was working on Recipe.Site full-time, while he had a few hours a week at most, which created tension. But I lacked the perspective to recognize the issues, and was too excited to acknowledge them even if I had.

V1

Conor had a clear idea of how to build a lightweight backend/CMS using Sanity, that could support an ingredient database and structured data that allowed for recipe scaling, conversions, dynamic inline ingredient references, etc. We’d essentially get a recipe editor UI for free, though it would be a jerry-rigged version of Sanity’s default interface, and clunky as a result (more evidence that typical CMSs can’t accommodate structured recipe writing).

I decided to handle all the recipe input myself for now (rather than let Alice do it), both to learn about recipes — how they are structured, their nuances and edge cases, etc — and to use my experience of a bad recipe editor to inform future designs for a good one. I populated the ingredients database, added all of Alice’s recipes, and a few weeks later we launched her site.

Meanwhile, I had been casually looking for other recipe creators. I found Jerumai on Instagram (he was posting mostly food photos and recipes at the time) and emailed him. He had a few strong opinions about the design of his site (colors, fonts, lowercase text, homepage layout, etc.), which led to conversations about customization. This is a common question for CMS/website building/blogging software: how much of the default design can someone modify? It’s why templates and themes exist. My intuition was that there would be a few common elements most people would want to customize (fonts, colors, logo, border radius of images), and a long tail of unique requests (header height, border thickness, recipe title font size, etc.). I felt that more customization was better than less, which tied into my bias towards tool/infrastructure over platform/container. Every creator signing up would already have social media profiles, and a customizable website is a nice contrast — it’s like a house you buy and make your own, versus an apartment you rent. I knew that more customization meant more maintenance, but I saw a lot of value in offering more customization than average and figuring out an architecture that would keep it sustainable.

After Jerumai posted about his site on Instagram, people started reaching out. Over the next few months, five more creators signed up and launched sites:

• Citrus and Roots
• Chung Eats
• Dr Matt Cooks
• Fifty First Tastes
• Krizkitch

Ryan

One of the creators who reached out was Ryan Nordheimer. Unlike the other creators, Ryan wanted a WordPress site with ads — exactly what I hoped Recipe.Site would replace. I gave him my take on why subscriptions made more sense than ads, but he was clear about what he wanted. I realized that building and maintaining a WordPress site for Ryan would then allow me to speak from firsthand experience when creators asked about WordPress and Raptive/ads. Additionally, compared to other creators I had met, Ryan was very organized and had a structured growth plan. So I figured it was a good relationship to develop (it was — I eventually migrated him off WordPress to Recipe.Site, he became the first creator with a Raptive/ad-enabled site on the platform, and the relationship has paid off in many other ways).

This experience got at the perennial question: am I building a design and development agency for food media websites, or a tool people use on their own? I leaned towards the latter, but also felt drawn to things on the periphery. In part this was because it’s hard to say no to any opportunity (and a little extra cash) at the very beginning of something. In part it was because I was realizing relationships would be critical to this business. And in part, stretching towards the right opportunities served as R&D — exploring things the business might one day encompass.

Year 3 (November 2024–November 2025)

Year 3 was largely “heads down,” as people in tech say. I had a clear sense of what I wanted the product to be, and I knew what I needed to do to grow the business. The two main objectives were:

1. Get off Sanity: build a backend and a recipe editor
2. Find more creators who wanted websites

Sanity -> custom app

This was a decent amount of work, but relatively straightforward. Drawing on my experience inputting recipes into Sanity, I sketched a simple recipe editor. Most of the interactions had been solved by existing apps (I referenced Notion a lot), and for the tricker bits I built little prototypes to feel it out. Conor built the interface and wired it up to write to Sanity. Next came a backend, then connecting the editor to it. There were some interesting technical challenges along the way, but I’ll leave that writeup to Conor.

The recipe editor was (and remains) the foundation of the entire app. It is the primary differentiator: no other recipe editor is as well designed or nearly as fully featured (capturing nuances like divided ingredients and baker’s percentages). Structured recipe data also enables the most valuable features for readers/cooks: inline ingredient references that scale with the recipe, recipe summaries, ingredient substitutions, unit conversions, advanced sorting and filtering, etc. We also built some light editorial and customization features: recipe collections, adding featured recipes to your homepage, changing your site name, adding an about page, creating a linktree-like page, etc.

The professionals

All the creators using Recipe.Site so far were doing recipe content on the side. The next group were professionals (doing food stuff full-time). Their website needs weren’t necessarily different, but they exposed me to other parts of the food world (recipe developing, testing, and styling, running restaurants, writing cookbooks, catering, events, etc.). Those creators/sites were:

• https://www.manololopez.com/
• https://www.cesar.nyc/
• https://www.kenjcooks.com/

Kenji was about to publish his first cookbook, which presented opportunities beyond a recipe website. First we built a simple landing page for his cookbook, with UK and US bookseller links and a way to flag promotions.

Next, we created a digital preview of the cookbook, which we hoped could replace the standard PDF sample publishers send to journalists and reviewers. We also discussed including an enhanced digital version of his cookbook with every purchase of a physical copy, though by then everything was signed and sealed, so it seemed effectively out of the question.

First subscription site

Each day I stared longingly across the field at Substack. I was envious of their community; so many talented food people were using, and seemed to love, Substack. And I was frustrated: all these great recipes, while thankfully not plastered with ads, lacked proper formatting, utilities, search and filtering, and basic quality-of-life features like keeping your screen awake while cooking. Substack gave recipe creators a new business paradigm (subscriptions instead of ads), but didn’t go the final mile to present their recipes properly. Two steps forward, one step back. I knew my recipe experience was better than Substack’s, and I was eager to find a creator who wanted to start a recipe subscription site and demonstrate an alternative.

Somewhat out of the blue my second cousin connected me to someone who wanted to start a recipe subscription site with a few friends. They knew about Substack but liked the idea of a smaller platform with more hands-on support. A few months later we launched Our Recipe Books and slowly but surely they started getting paid subscribers (they’ve since made all the recipes free). It was a small subscription business, but a subscription business nonetheless, and a perfect, forgiving, low-stakes partnership while we were still working out the Stripe subscription wiring.

Joshua Weissman

In Spring 2025, I was introduced to Joshua Weissman. I knew he had an partially broken Wix site. Lawrence, his Head of Production, told me they had spent months migrating their content to Webflow and were close to launching, but needed a little help. He was curious about Recipe.Site but mainly wanted help getting the new Webflow site over the line. Back, once again, to the perennial question. I knew the relationship would be valuable (I already liked Lawrence) and I was familiar with Webflow, so the answer was obvious enough.

After I signed a contract for a small amount of work, Conor nudged me to pitch them on switching to our platform. I didn’t want it to feel like a bait-and-switch, so I made sure to finish all of the work they asked for, then spent the weekend creating a copy of their new Webflow site on our platform. I thought the biggest sticking point would be the recipe migration (they had just spent months migrating and reviewing their content, and would have to do it all again), so I tasked Leo (our summer intern) with building an LLM-based recipe parser to import 700+ of Josh’s recipes. They’d need to be reviewed and cleaned up by a person, so I started looking into hiring a data entry contractor.

Conor and I met with Lawrence to show him the changes to the Webflow site, and once he was happy with where things stood, Conor pitched the switch. We showed him the site, all of the additional functionality they would get out of the box (scaling, conversions, searching, filtering, comments, etc.), talked about the value of Josh doing something well beyond the satus quo, and explained the recipe migration workflow. He was direct with us: they wanted to launch on Webflow asap, and might consider switching later. No promises, no expectations.

I decided to take a small gamble: hire Alexandria (the data-entry person), migrate and review all of the recipes (using Leo’s recipe parser), keep improving the Recipe.Site version of his site, and pitch again once everything was in place. They ultimately decided switching to a new platform with a small team was too risky. But I’ve since many happy hours maintaining and improving their Webflow site (which I can at least say is the best recipe website built on Webflow), my relationship with Lawrence is wonderful, and, as of summer 2026, there are embedded Recipe.Site components that handle comments and ratings (which Webflow doesn’t offer natively). Perhaps the biggest silver lining: Alexandria is now full-time and leads operations.

A community cookbook

Most of the software I’ve seen in the food/recipe space doesn’t have much of a soul. Look at some recent startups: Dishy, ReciMe, Provecho, Crouton, Pepper. They all look lifeless and the same. My hunch is that this is because the people building those products don’t have a personal connection with what they’re building. They probably had a general interest in software, put a bunch of ideas on a whiteboard, ran the numbers, and for whatever reason picked the recipe idea. Which isn’t to say that they don’t cook — I would guess they all do. But there is a difference between having a general interest in an activity and being passionate, even obsessive, about it. I think to create something unique and valuable in this space requires being equally obsessed with recipes/cooking/food, design, and engineering.

Before I started Recipe.Site I was passionate about cooking, and over the first three years became obsessed with how recipes are written and read. I have also become interested more broadly in food culture, systems, and communities. While some of that passion came across inherently, I started thinking about things I could do to further explore and express it, both because it helps build my intrinsic love and motivation to continue, and because I think it builds trust with the outside world and helps attract the right kinds of collaborators and partners.

Many ideas had come to mind (writing about food, interviewing people, sharing my pizza recipes, etc.), but I decided to start with a community cookbook. It felt like a nice balance of exploring the food world in a new way, getting to meet some people, using and promoting a new feature of the product, and demonstrating that I actually care about this stuff and the people doing it. I asked Steph Lau (who I had worked with previously) if she would be interested in curating/editing it. She came up with a theme, got a wonderful group of people to contribute recipes, and decided that it should be a fundraiser for World Central Kitchen. At that point I only had a prototype of a digital cookbook that I had made for Kenji, so Conor formalized the concept of a cookbook in the app, and together we made a bunch of refinements (table of contents, cover page, pagination, previous/next page buttons, buy button, etc.). This is the community cookbook we published (named, somewhat painstakingly, by Conor): Mixtures: Vol. 1

Farewell to Conor

By October I had to face the reality that Conor and I were no longer working well together, which was hard for me. It took a while to find the words, then a few more months to work out a separation agreement, but eventually we parted ways.

Reflections

Two things stand out to me when I look back on the first three years:

1. Starting something from scratch requires a lot of patience and persistence. One of the defining feelings was that I was trying to breathe life into some fledgling being who was on the verge of dying. There’s a gentleness to it, which is the patience part, and also a steadfastness, which is the persistence part.
2. Before I started I had some fantasy that once I built the thing and put it out in the world people would simply find it, use it, and love it. I could remain anonymous, behind the scenes, tending to my little digital garden. Instead, the business is almost entirely about relationships. In part a rationalization of not having a plan, I’ve started to believe that plans are far less relevant than relationships. I like the idea of pointing your efforts in a specific direction, and then using the relationships that you form to emergently guide you forward.

Background

Inclinations

From 2007 to 2011, I ran Create Build Destroy (a small clothing company) with my cousin and two friends. Building something from scratch, being responsible for many different aspects of it, and growing through constant challenges (including learning some graphic design and coding), was one of the most fun experiences I’ve had. My dad had been a successful entrepreneur my whole life, so there was also a general culture, romanticization, and maybe even expectation of entrepreneurship in my family.

From 2015 to 2018, while living in San Francisco, three food/cooking related things happened:

1. I started cooking. I found two recipes I loved on New York Times Cooking (Three Cup Chicken and Longevity Noodles), and cooked them over and over. The excitement of being able to cook anything at all was exaggerated by the fact that these two dishes were sort of “restaurant food” in my mind. I guess I had a sort of vague mental distinction between home food and restaurant food. Until I started cooking those two dishes, I hadn’t realized you could make “restaurant food” at home. I also took a few cooking classes at SF Cooking School, and started cooking with two friends of mine (Veronica Borges and Alex Gibbons).
2. I read Michael Pollan’s The Omnivore’s Dilemma, visited Full Belly Farm (northern California) and Polyface Farm (Virginia), and started to get more interested in food and agriculture generally
3. At AltSchool, where I worked as a designer, the take-home design challenge we gave to applicants was to redesign some aspect(s) of the recipe discovery, grocery shopping, and cooking experience. Though I don’t remember many specifics, I remember being intrigued by a lot of the ideas. And I remember Veronica (who also worked at AltSchool) often saying that we should actually build something in the recipe/food space.

Learning to code

Since I started designing software, I had always idolized people who could design and code. I remember hanging out with Greg Passuntino, Jonathan Greacen, Eli Rousso, Chris Parcel, and Alex Gibbons, watching them build things they had designed, mesmerized, asking them questions, hoping for some kind of osmosis. I was fascinated, but coding felt generally unattainable to me…sort of like this binary thing: there were people who coded and people who didn’t. The most I was doing was hacking together custom WordPress themes and building static HTML/CSS sites with minimal jQuery animations/interactivity, which didn’t really count for anything in my mind.

In 2017, I turned the corner. I took an in-person JavaScript course where I built my first “application”: a photography portfolio website connected to a database, where you could add and remove images. It was an incredible moment. Similar to cooking Three Cup Chicken, I had crossed some kind of invisible threshold. Excitement, confidence, etc. The next course I took was an online React course. For the class projects I built a cooking timer and an ingredient substitution app. I vaguely remember thinking at the time that those pieces were going to eventually fit into some larger whole (especially the ingredient substitution app — by itself it would never be very discoverable or useful, but that it was worthwhile because I’d eventually use it in some bigger thing).

The design of recipes on the internet was originally published in Bootcamp on Medium, where people are continuing the conversation by highlighting and responding to this story.