Vita

Inspiration

In the whirlwind of our fast-paced lives, the simple pleasure of a home-cooked meal often gets lost. Amid the chaos of work, errands, and fleeting spare moments, many of us succumb to the lure of fast food and convenient takeaway options. According to the National Center for Health Statistics, an alarming 36.6% of American adults are daily consumers of fast food. This trend started as a vague observation, until one day, it became a personal reality for a close friend. Our friend, an incredibly busy and ambitious woman, was constantly juggling schoolwork, a part-time retail job, and a bustling personal life. On top of it all, she had to contend with various dietary restrictions, particularly the need to avoid gluten. With her love for food and a desire to maintain a healthy lifestyle, she often found herself planning to cook at home but always came up against the same obstacles: limited time and the daunting task of meal planning and research. Existing recipe platforms did provide a starting point, but they left much to be desired in terms of personalization and dietary accommodation. Recipes lacked the flexibility she needed, and a limited database meant she ended up cycling through the same meals, despite her adventurous, multi-cuisine palate. As we watched her struggle, a question began to form: Could we bridge the gap between the convenience of fast food and the benefits of home-cooked meals? The answer came in the form of Vita - an AI-powered personalized cooking companion designed to make home cooking a viable, enjoyable option again. Vita is more than just a tool - it's a way to promote healthier living. Home-cooked meals aren't just healthier; they're also a smart economic decision. They cut down on excessive spending on fast food and reduce our reliance on the convenient but often unhealthy options available for takeaway. Our inspiration for Vita comes from personal experience, the desire to solve a genuine problem, and the belief that everyone deserves the joy and benefits of a home-cooked meal. By harnessing the power of artificial intelligence, we aim to bring back the charm of home cooking, one personalized recipe at a time.

How We Built Vita

In the quest to build Vita, we began with detailed planning and prototyping. We believe that a successful application is one that's meticulously designed around the user's needs. Thus, we adopted a user-centric approach and laid out our application's blueprint on Figma, a collaborative interface design tool. The primary focus during this planning stage was four essential features that make Vita Vita: survey preferences, recipe generation, grocery list creation, and social interaction. We created a detailed wireframe to envision the user flow and interaction link. For the survey preference section, we brainstormed an extensive list of dietary requirements, food preferences, and taste factors. We designed the UI in such a way that users can easily input their specifics. Next, we prototyped the recipe generation feature. Here, the goal was to ensure that users receive custom recipes, tailored according to the inputs given in the survey preferences section. To add more value, we envisioned an AI-powered engine that could generate these personalized recipes. We then turned our attention to the grocery list creation feature. Vita wouldn't just recommend recipes but also intelligently generate a list of ingredients needed for these recipes. We designed this feature to streamline the grocery shopping process, freeing users from the hassles of manual list preparation. Finally, we designed the social interaction feature, a concept aimed at fostering a sense of community among Vita users. Our wireframes focused on enabling users to share their culinary journeys, swap tried-and-true recipes, and learn from each other's experiences. We prioritized building a section where users could explore similar, tested recipes from the community, making their culinary exploration more reliable and engaging. Once the prototypes were complete, we transitioned to the building phase. We chose to develop a Next.js web app, using a toolkit comprising of create-t3-app, Tailwind, TRPC, and TypeScript. Our key ingredient was the integration of OpenAI's powerful language model (LLM), which we incorporated through AWS Lambda functions. This AI-driven model serves as the core of Vita's recipe generation, creating unique recipes based on user preferences. To further refine the recipe recommendations, we utilized Pinecone and Cohere text embeddings.

For our backend infrastructure, we selected Supabase, a reliable serverless solution. Supabase served as the backbone for our database requirements and authentication processes. This decision significantly reduced the complexity and time required for backend development. Finally, to ensure the seamless functioning of our AI-driven features, we needed a Python runtime. We found a perfect fit in Vercel's serverless environment. It provided the robust support our application needed and complemented our technology stack effectively. From sketching the first prototypes on Figma to writing the final line of code, every stage of building Vita was carried out with the user's needs at the forefront. Through a systematic and iterative approach, we breathed life into our idea and saw Vita transform from a mere concept to a functional application.

Challenges Along the Way

Every ambitious endeavor encounters its unique set of hurdles, and our journey with Vita was no exception. With time ticking away, we found ourselves in a constant race against the clock, striving to convert our ideas into tangible features. The primary challenge was defining the scope of our Minimum Viable Product (MVP) amidst numerous potential features. We faced the dilemma of what to prioritize - do we focus on building a multitude of features or channel our energies towards refining a few core functionalities? It was a classic case of quality versus quantity, and we chose to focus on quality, emphasizing the refinement of our key features - the personalized recipe generation and user experience. We believed these were the elements that would genuinely make a difference in our users' lives, providing them with a seamless and value-added experience. Additionally, there was the challenge of tailoring our app's layout specifically for Devpost. We understood that the user interface (UI) could greatly influence the overall user experience (UX). So, instead of cramming in more features, we prioritized enhancing the navigational flow and usability. This led us to replace traditional app buttons with a more intuitive and minimalist hamburger menu and reduce the need for horizontal scrolling. In overcoming these challenges, we didn't just build an app - we gained invaluable insights about prioritization, user-centric design, and the art of balancing quality and quantity. These lessons shaped Vita's development and will continue to guide us in our future endeavors.

Looking Forward

We envision a future where Vita is more than just a recipe app. We plan to integrate it with delivery services, expand its social features, and provide even more personalized options. Potential additions include a grocery receipt scanner using OCR technology and a comprehensive nutrition planner that extends beyond the home kitchen. Our dream is to make healthy meals - especifically at-home ones - accessible, enjoyable, and effortless for everyone. Vita is a step towards realizing that dream. We can't wait to see what delicious journey awaits our users!

Built With

• cohere
• duckduckgo
• golang
• lambda
• nextjs
• openai
• pinecone
• python
• supabse
• typescipt
• vercel
• verclcron