A mobile platform that helps its users to browse and shop an exclusive selection of hand-picked wines, beers, and spirits, discover personalized recommendations from the inbuilt chatbot, easily sort by type, country, grape, or price, and personalize delivery options during the checkout
In this project, our team has helped to design, develop, launch, and support a custom mobile user interface that has replaced the old no-code application. Our team was responsible for new branding, UI/UX for mobile application mock-ups, cross-platform iOS and Android app development using React Native, connection to API endpoints and front-end business logic, Stripe payment gateway integration, error handling, testing, promo materials for App Store and Google Play, and help with builds submission to review and distribution. Our project manager, UI/UX designer, Mobile developer, and QA engineer in strong conjunction with Swizzle CEO Capucine Codron transmitting business requirements and project guidelines, and CTO Guillaume Jacomin, who was building API and back-end integrations using Python Django framework.
| Industry: | Ecommerce, Food Retail |
| Platforms: | Mobile |
| Role: | Software developer, IT consultant |
| Tech Stack: | Python Django, PostgreSQL, React Native, Google Firebase, Twilio, SendGrid, Stripe, Google Analytics, Meta Pixel, AWS |
Swizzle CEO Capucine Codron has reached out to our company with a request to build a custom React Native application that should replace the existing one. Swizzle CTO Guillaume Jacomin was building back-end API using Python, and some API endpoints have already been documented. Additionally, we had an old application as a reference and brief information about the business and required functionality. Initially, the no-code mobile application they had was enough for the start-up. But as their business grew, they needed a more advanced back-end and user interface, and it was impossible to resolve just with a no-code framework. Thus, they requested several estimation options for just web or mobile apps or both. Based on our rich experience in food tech and retail software development, we quickly collected the required details, proposed several versions of the ballpark estimates for several options, shared our suggestions, and agreed on the design and development of the cross-platform mobile application on a T&M basis with an eye on budget and timeframes proposed in recent estimates.
In general, there were no overwhelming challenges. Both project team members and their Powercode team leads are deeply experienced in areas of e-commerce, food-tech, retail, and delivery, are familiar with different payment gateway integrations, and loyalty program implementations. Both traditional, marketplace, and recurring transactions and any complex monetization logic are not an issue at all. So in most cases, we just had minor routine organizational and communicational problems, which were monitored and resolved during the day under the watchful supervision of our technical project manager. For instance, our mobile developer worked separately from Guillaume, who developed the back-end, and in different time slots, so it was really important to sync team daily goals, clearly report on progress and problems, and resolve or avoid any miscommunications that could cause extra effort for both developers.
Regarding functionality, both the Wine browsing screen and Chatbot had rather complex implementations. The Wine browsing screen has a lot of selecting filters, a search bar, and a slider for selecting a price range. The set of all screen filters is synchronized with the filters in the search window. Filters are divided into two types: chips, and sliders. Each type has a different logic for building a search link. Moreover, screen content flexibility also adds complexity. Since the content of the application is managed by the admin, we can have different sets of blocks for the campaign screens and the main browsing screen.
Regarding the Chatbot, indeed, it’s not just a bot. All content is initially obtained from the back end, and we simulate the communication with the user. It has a large amount of logic for simulating a set of messages and handling the user’s response. Each user’s response is converted to a search query, from which we compile a link with a list of products, individually recommended to the user.
It was a real pleasure to build and release this product to market. In close coordination with Capucine and Guillaume and due to their energy and passion our team managed to resolve all the technical blockers, cope with all the product challenges, and deliver worthy and unique solutions with joy and enthusiasm. It was an exciting journey and we are happy that it’s not the end because we are currently helping with tech support and updates related to both back-end and front-end tasks.
