Reny: AI cooking coach

Reny is an AI cooking coach that cooks with you in real time - guiding every step, answering questions, checking your food with the camera, rescuing mistakes, and helping you improve with every meal.

Comment

Inspiration

Most cooking apps are great at telling you what to cook, but they stop being useful the moment you actually start cooking.

A recipe can tell you to “sear until golden,” “reduce the sauce,” or “cook until done” — but those are exactly the moments when many people become unsure.

Reny also started from something very personal.

I’m building the project together with my mom, who is currently growing her own cooking channel on social media. Through that, we see first-hand how much people enjoy discovering recipes and watching cooking content — but also how big the gap is between watching a recipe and confidently making it yourself.

People may save a recipe they love, but when they actually stand in the kitchen, completely different questions come up:

  • Is this browned enough?
  • Should I flip it already?
  • Is the sauce too thick?
  • Can I substitute this ingredient?
  • How do I fix this if something goes wrong?

That made us ask:

What if a recipe could actually stay with you while you cook?

We wanted to build something that feels less like a recipe database and more like having an experienced cook next to you.

That became Reny: an AI cooking coach that cooks with you in real time.

The combination also felt natural for us: my mom creates the food and cooking content, while I build the product and technology behind Reny.

Our goal is to connect those two worlds — discovering food through social media and actually being able to cook it successfully yourself.

What Reny does

Reny turns almost any recipe into an interactive cooking session.

Users can bring recipes into Reny from social media, websites, screenshots, photos, or even handwritten recipe notes. Reny turns them into structured recipes that can be saved and cooked with the same guided experience as recipes already available in the app.

Instead of simply showing a list of instructions, Reny guides the user through the cooking process step by step and understands the context of the current session.

Users can:

  • Import recipes from social media, websites, screenshots, photos, and handwritten notes
  • Follow recipes in a focused step-by-step cooking mode
  • Use smart timers and contextual timing guidance
  • Ask questions by voice or text while they cook
  • Show Reny their food with the camera and get visual feedback
  • Get contextual help when something goes wrong
  • Receive guidance adapted to their cooking level and goals
  • Receive a Cooking Score after finishing a dish
  • Earn XP and improve individual cooking skills over time
  • Take part in creator challenges and try to beat previous scores

For example, instead of only reading “cook the meat until browned,” Reny can tell you when to wait a little longer, let you check the food with the camera, and help you decide when to continue.

If something goes wrong, the user does not have to leave the recipe, search Google, or open a separate AI chat.

They can simply ask Reny or show what is happening.

The goal is to make cooking feel more interactive, less stressful, and ultimately help users become more confident and capable cooks over time.

How we built it

Reny is built as a mobile-first app using React Native and Expo.

We designed the cooking experience around a state-based session rather than a traditional recipe page.

Reny always knows:

  • which recipe the user is cooking
  • which step they are currently on
  • what has happened earlier in the session
  • which timers are active
  • the user’s cooking level and goals

AI is integrated directly into that context.

That was important to us because we did not want Reny to feel like a generic chatbot added on top of a recipe app.

The camera experience lets users visually check food or show Reny a problem.

Voice and text interactions let users ask questions without leaving the cooking flow.

We also built progression into the experience through:

  • Cooking Scores
  • XP
  • individual cooking skills
  • level progression
  • creator challenges

This turns cooking into something that users can improve at over time instead of simply completing individual recipes.

For monetization, we integrated RevenueCat to manage subscriptions, trials, offerings, and Premium entitlements.

Reny has a free experience, while Premium unlocks the complete AI cooking coach.

Using RevenueCat allowed us to keep subscription logic separate from the rest of the product and made it much easier to manage access to Premium features without building our own subscription infrastructure.

Challenges we faced

One of the biggest challenges was deciding when the AI should help and when it should stay out of the way.

Cooking is very different from a normal chat interaction.

Nobody wants to constantly stop cooking, type a question, wait for a response, and then go back to the pan.

The experience had to feel fast, contextual, and lightweight.

We also had to distinguish carefully between different types of intelligence.

For example:

  • timing guidance should not pretend that Reny can visually see the food
  • visual feedback should only make visual claims after the user actually shows the food
  • voice questions, camera checks, and troubleshooting should feel like separate capabilities while still being part of one continuous cooking session

Another challenge was AI cost and latency.

Continuous video analysis would be expensive, slow, and unnecessary for most cooking situations.

Instead, we designed the system around deliberate camera checks and event-driven interactions.

That gives users the feeling of having an AI cooking companion without continuously streaming video in the background.

We also spent a lot of time thinking about onboarding.

The value of Reny is difficult to explain with a simple feature list.

The product only really makes sense once someone experiences what it feels like to cook with it.

So instead of immediately showing a paywall, the onboarding includes a short simulated cooking session where users experience timing guidance, camera checks, voice interaction, troubleshooting, scoring, and challenges before being introduced to Premium.

What we learned

The biggest lesson was that an AI product becomes much more useful when the model has context and a clear role.

A generic AI assistant can answer cooking questions.

But an assistant that knows:

  • what you are cooking
  • which step you are on
  • how long something has been cooking
  • what your cooking level is
  • what your goals are
  • what problem you are currently facing

can give much more useful and relevant help.

We also learned that the best AI interactions are often very small.

A short message like:

“Wait another 30 seconds before flipping.”

can be much more useful in the kitchen than a long AI-generated explanation.

That changed how we designed the product.

Instead of maximizing how much the AI says, we started optimizing for timing, relevance, and context.

We also learned that AI should not try to appear more capable than it actually is.

If Reny has not seen the food, it should not claim to know exactly how brown it is.

If it only has timing information, it should give timing-based guidance.

That distinction became important for making the experience feel trustworthy.

What's next

Reny is just getting started.

One of the things we are most excited about is connecting Reny even more closely with the cooking content my mom creates.

Instead of a social media recipe ending when the video ends, we want viewers to be able to open the same dish in Reny, cook it step by step, ask questions, get visual feedback, and eventually compare their own Cooking Score.

That creates a much more interesting loop:

Discover a dish → cook it with Reny → improve → take on a challenge → discover the next dish.

We also want to:

  • make Reny increasingly proactive and personalized
  • improve visual food understanding
  • expand cooking skills and progression
  • build more creator challenges
  • expand recipe discovery
  • connect social cooking content more directly with the in-app cooking experience

Over time, we want Reny to feel less like an app you open for a recipe and more like a cooking companion that actually helps you become better in the kitchen.

Our long-term vision is simple:

Recipes tell you what to do. Reny helps you actually do it.

Built With

Share this project:

Updates

Submission history