ABU PICK 4 ME Ai

ABU PICK 4 ME AI is a campus food service that suggests 3-4 meals from nearby shops based on your orders, budget, and class schedule, with tap-to-order, group orders, and smart delivery routing.

Comment

Inspiration

What Inspired Me: The project was inspired by the daily struggle of ABU students trying to order food between lectures. Too many menus, long wait times, and decision fatigue made eating on campus inefficient. I wanted to create something like Haru and Gaku’s “yorinoku” service—an AI that picks for you—tailored specifically for ABU’s campus life and student budgets.

What it does

ABU PICK 4 ME AI is a campus food service that suggests 3-4 meals from nearby shops based on your orders, budget, and class schedule, with tap-to-order, group orders, and smart delivery routing.

How we built it

How I Built It: I designed ABU PICK 4 ME AI as a mobile-first service with three core layers:

  1. User Profiling: Captures order history, budget range, and typical free periods.
  2. Recommendation Engine: Filters nearby vendors and surfaces 3–4 meals that match taste, price, and delivery window.
  3. Logistics Layer: Groups orders from the same hostel or building to optimize rider routes and reduce delivery costs.
    The brand identity uses a green shield logo with a fork-and-spoon pin to keep it rooted in ABU’s identity while staying clean for WhatsApp and flyers. ## Challenges we ran into Challenges Faced: The main challenge was data sparsity—new users have no order history, so cold-start recommendations had to rely on campus-wide trends and time-based patterns. Coordinating group orders and real-time rider routing also required careful logic to avoid delays. Finally, keeping the experience fast and low-friction on WhatsApp, where most students operate, meant stripping away unnecessary steps. ## Accomplishments that we're proud of Accomplishments that we're proud of:

  4. Clear Value Proposition for ABU Students We defined a focused problem: decision fatigue and time constraints between classes. The “Pick 4 me” concept gives students 3-4 relevant meal options in seconds instead of endless scrolling.

  5. Campus-Specific Brand Identity Created a green shield logo with a fork-and-spoon pin that ties directly to ABU’s identity while staying clean and recognizable on WhatsApp, flyers, and rider gear.

  6. Practical AI Logic Design Mapped out a recommendation system that uses order history, budget, time between classes, and delivery window. It’s designed to work even with limited data by leveraging campus-wide trends for new users.

  7. Logistics-Aware Approach Built group ordering and batch routing into the core concept. This reduces delivery costs and makes the service viable for low-margin student pricing.

  8. Mobile-First, Low-Friction UX Prioritized WhatsApp as the main interface so students don’t need to download another app. The flow is tap-to-order, quick replies, and minimal steps.

These milestones give us a working foundation to move from concept to a real pilot on campus.

What we learned

What I Learned Building this taught me how much context matters in recommendation systems. It’s not just about food preferences; time between classes, group ordering behavior, and delivery routing all shape a useful experience. I also learned to balance simplicity in the UI with enough intelligence in the backend to make the “pick for me” feature actually feel helpful.

What's next for ABU PICK 4 ME Ai

What’s Next for ABU PICK 4 ME AI:

  1. Pilot Launch & Data Collection Start with 2-3 halls and 10-15 vendors around ABU Zaria. The goal is to collect real order data so the AI stops relying on campus-wide averages and starts learning individual and hostel-level patterns.

  2. Smarter Recommendations Add time-aware and context-aware picks:

  3. Pre-class rush: Fast meals under 10 min delivery

  4. Group mode: Auto-suggest meals that split well for 3-5 people

  5. Budget mode: Prioritize ₦500-₦1500 options during month-end

  6. Rider & Vendor Tools
    Build a simple dashboard for vendors to manage menus and for riders to see batched routes. Better logistics = lower delivery cost and faster drops.

  7. WhatsApp Integration Keep ordering in WhatsApp since that’s where students already are. Use quick-reply buttons for “Reorder”, “Pick 4 me”, and “Group order” to keep it low-friction.

  8. Feedback Loop Add a 1-tap rating after delivery. That feedback directly retrains the recommendation model so bad picks get filtered out fast.

The next 30 days are about getting 200+ real orders through the system. Once the data is there, the AI actually becomes useful instead of just a concept.

Built With

  • fluterwave
  • python
  • whatsappbusiness
Share this project:

Updates