Stealth FoodTech Start-up
A full-stack experience: smart oven, chef-prepared meals, and an app that brings it all together. Designed to redefine how busy people eat at home.
Idea & problem
Busy people want to eat well at home — but existing food delivery solutions are either too slow, too expensive, or too limited in nutrition.
We set out to create a smarter experience: high-quality chef-made meals, delivered frozen, reheated in minutes with a smart oven, and fully controlled via a mobile app.
Product
Food + Smart device + Mobile app + Website + Delivery
Goal 2024-2025
Design and launch the core product experience from scratch: smart food ecosystem with mobile-first ordering and smart oven integration.
Investment
Angel round — €50M
Timeline
July 2024–April 2025
Growth team
but overall our team is about
100 people
VP Growth
Yulia Voronina
Product Lead
Artem Ivanov
Creative Director
Vincent Riezouw
Senior Product Designer
Kris Barilnik
01. MVP in 2 months
As the growth team, we designed, developed and launched the first version of the app in under 2 months — covering the full journey from onboarding to payment and tracking. Built on React Native, which allowed us to move fast across platforms, despite certain design constraints.
Since there were no direct analogues on the market, we based the product on the founder’s vision and deep research into delivery services and smart kitchen ecosystems. In parallel, the oven was being developed by a separate team, chefs were preparing meals for shock-freezing, and the ops team built out logistics and customer care. We stayed tightly aligned to ensure everything worked as one system.
As part of the growth team, we also developed the first version of the brand and customer communication — including food pod labels, packaging, and brand language across touchpoints.
02. Fast iteration & live experiments
The early cycles were focused on delivery speed and operational readiness — ideal for a fast-moving startup.
Ahead of our first internal test, we ran rapid-fire experiments to validate key parts of the user journey. We tested multiple approaches to presenting nutritional info, portion sizes, and visual clarity of food, explored different structures for time slot and date selection, and optimised how users add, edit and confirm their delivery address.
We also experimented with order tracking visibility and messaging — shaping how and when users receive updates. Operational feedback loops were tight, allowing us to ship and iterate on improvements every week.
By this point, the app already covered the full end-to-end flow needed for the first internal testing, including:
User authentication and closed-access onboarding
Guided onboarding journey for first-time users
Menu browsing with filtering and tagging
Detailed food view
Add-to-basket and checkout flow
Address management
Delivery slot and date selection
Payment via Stripe integration
Real-time order tracking
Full order history with status breakdown
One of our core product hypotheses was that if we helped users eat in a balanced way, they would naturally choose proteins, vegetables and carbs in healthy proportions. That’s why both in the app and on the packaging we clearly categorised food by nutritional role — making it easier to build balanced meals.
From the business side, it was important to consistently promote protein dishes, as they were the most profitable — and we adjusted hierarchy and placement to support this.
03. First test:
connect everything
Initial test focused on stitching together all parts of the system to validate the end-to-end mechanic and uncover grey zones — areas where the flow broke down or user expectations were not met: food selection, ordering, and delivery loop. Integration with the smart oven was not tested at this stage, as the hardware prototype was still in development.
Test scenario: log in, add delivery address, select meals, choose delivery date and time, place an order, and receive the delivery.
Reworked several flows based on test outcomes:
Redesigned portion size selection to improve clarity
Refined the number and content of system emails
Optimised slot & time selection for better cost-per-order
Introduced rules to hide unprofitable delivery options while keeping the design predictable
Made address input more flexible for dense urban contexts like London: added address type selection, extended delivery option information, enabled comments per delivery type, and allowed users to upload photos and notes with their address
After implementing these changes, we were ready to move forward with the next round of testing.
04. Second test: polish the system
The second internal test focused on verifying that all fixes were in place and the full order journey now worked seamlessly. Every step — from browsing meals to receiving delivery — was reviewed and refined.
This time, each participant was assigned a tailored test scenario with specific actions and outcomes:
Delayed deliveries
Delivery to the wrong address
Damaged deliveries
Undelivered packages
Missing or wrong items delivered
Technical issue with app
Insufficient stock
These variations helped us observe how well the app and customer care flows handled edge cases in a live environment.
Based on feedback from the second test, we revisited how meals were displayed and selected:
As variety increased, food discovery became harder — we introduced collections, filters and tags, and enabled food discovery through tappable tags and ingredients within each meal view
Added an image gallery inside the dish screen with flexible layout options (top, side, full bleed). Later considered 3D models of meals generated via ready-made tools
Simplified the design and improved readability of nutritional and ingredient information
Introduced food pairing suggestions to help users build balanced orders
By this stage, our branding had evolved — we updated the app’s background to a soft grey to visually align with the oven and other touchpoints.
After a number of testing sessions, we also selected a new plate design that became the standard for all food photography — helping unify the visual style across app, packaging, and marketing.
In parallel, we started integrating the oven prototype with the mobile app, and launched a supporting website to explain the ecosystem — what the product is, how it works, and why it’s different. The site became a key part of the conversion funnel.
We also added a review feature in the app to collect feedback on both the meals and the delivery experience.
05. Third test: simulate real life
By the time of the third test, the first batch of oven prototypes was ready and distributed to users from our closed test group. The goal was to validate the full cycle — from ordering and delivery to heating meals in the smart oven.
For this test, we designed and developed the oven integration and implemented all expected cooking events and error states. While we didn’t have time to fully test these scenarios end-to-end during the third test, all event handling and system logic were already implemented and ready within the app.
We introduced a layered interaction model: the top layer focused on food and delivery, while the bottom layer captured oven status and interactions. This helped users distinguish between system messages coming from the app vs. those from the oven itself.
We also categorised oven error types into two groups: blocking errors that stopped the heating process (highlighted in red), and non-blocking ones that allowed heating to continue (displayed in black).
In addition to that, we also:
Moved the app to the production environment and tested all flows in live conditions
Tested oven prototype with live cooking sessions and edge cases
Collected in-app feedback with animated review flow to close the loop
The third test was our longest — it ran for nearly two months. Over this period, we actively monitored results and continuously refined the product. Below are the core features we validated and integrated: Here are the key features we validated and shipped:
Added iOS Spotlight search to allow users to quickly find meals from the home screen
Introduced Wallet for better management of saved payment methods via Stripe and to track vouchers issued by customer care
Integrated Intercom chat support to streamline conversations between users and customer care
Introduced visibility of refunded items in the order history
Began updating the oven pairing flow to make the experience more guided and engaging
06. Full system launch & brand design
By April 2025, the full end-to-end product journey was in place — from selecting meals in the mobile app and purchasing the oven, to food delivery, heating in a functioning prototype, and reviewing the meal experience.
Alongside this, we developed a cohesive brand system that spanned every user touchpoint — mobile app, oven interface, packaging, emails, and marketing site. The visual identity was extended to internal tools and pitch decks, creating a consistent and recognisable experience across the entire ecosystem.
Not just an app. Not just an oven. A system — and it works.
