SirPrep

Dashboard
Recipes for each meal
Meal planning
Meal planning overview
Generated meal plan
Grocery list and alternatives
Macros and suggestions
Preference profile
Preference profile
Liked/avoided meals
Meal plans history

Inspiration

As students with part-time jobs and a motivation to build fun projects in our free time, we often found ourselves eating unhealthy food, spending too much money on snacks, and making unplanned grocery or takeaway trips. With SirPrep, we wanted to solve a problem we personally experienced: making healthy eating easier, cheaper, and less stressful for busy people.

What it does

SirPrep is a meal planner that learns your likes, dislikes, budget, and nutrition goals to help you eat healthier and plan smarter. It helps users create a full week of meals in just a few minutes, reducing unplanned grocery visits, food waste, and unnecessary spending.

Over time, SirPrep improves the planning experience by remembering user preferences, saved meals, dietary needs, and fitness goals, making each weekly plan more personalized and useful.

How we built it

SirPrep is a full-stack web application built on Vercel. SirPrep uses Amazon Aurora PostgreSQL-Compatible Edition because the product needs relational guarantees around household ownership, billing, credits, and calendar history, while still allowing for flexible AI-generated meal and recipe payloads as PostgreSQL JSONB. Aurora is the intermediate that connects the rest of the stack: Clerk provides the authenticated clerk_user_id used to scope every user-owned row, Vercel-hosted Next.js server components and server actions read and write through the server-only lib/db PostgreSQL layer, Gemini-generated plans and recipe cards are saved only after server normalization, and the Stripe webhook syncs normalized subscription state into the same user-owned data model. Aurora stores household's Planning Profile, saved weekly meal plans, liked or avoided meal memory, meal-plan credit usage, generated recipe cards, and subscription state, so the app can carry each household's planning history, preferences, and entitlements across sessions.

Challenges we ran into

One of the biggest challenges was making meal planning feel fast and simple while still collecting enough information to personalize the experience. It was difficult balancing flexible AI-generated suggestions with practical constraints like budget, ingredients, cooking time, and realistic student lifestyles. Moreover, we had to think carefully about how to structure user preferences, nutrition goals, and generated meal plans so the product can improve over time.

Another challenge was finding current prices for grocery store listings and offers as there is no public API available for use. Hence, we currently only offer price estimates.

Accomplishments that we're proud of

We are proud that SirPrep solves a real problem we personally face and turns it into a useful, polished product experience. We built a full-stack application with authentication, payments, database storage, and AI-powered meal planning working together. We personally started using SirPrep to plan our weekly meals and are very happy with the results. Our friends have also started using the app and we have begun running add campaigns and exploring marketing strategies to bring in further users to experience meal planning and its benefits.

Another purpose of ad campaigns is to test interest in SirPrep and understand which audiences respond best to the idea. These campaigns help us learn how people describe their meal planning problems, what messaging works, and which features users care about most. Our plan is to use these insights to improve the product, refine our positioning, and run more targeted campaigns as we prepare for launch and growth.

What we learned

We also learned a lot about integrating AI into a real product, handling user preferences, and designing an experience that feels helpful instead of overwhelming.

Most importantly, we learned that a good AI product is not just about generating content. It needs to understand the user, fit into their routine, and get better the more they use it.

What's next for SirPrep

Next, we want to make SirPrep more personalized and reliable by improving how it learns from user feedback, saved meals, skipped meals, and nutrition goals. We also plan to add smarter grocery lists, budget tracking, recipe discovery, and better support for different diets and fitness objectives.

We hope to eventually be able to provide users with realistic price points and offer-aware grocery lists. Furthermore we want to include nutritional goals of users in our app too: Building Muscle, Losing weight etc.

On the growth side, we plan to continue running ad campaigns, test different target groups such as students, young professionals, and fitness-focused users, and use the results to shape both the product roadmap and our launch strategy.