Most wellness apps track one thing. HivRa connects everything - your meals, workouts, habits, streaks, grocery planning, and AI-powered meal recommendations - into one intelligent system that learns from your routine, adapts to your activity, and helps you stay consistent every single day.
Wellness routines don't fail because people stop caring. They fail because the tools are scattered. Meal ideas live in saved Instagram posts. Grocery lists live in notes apps. Workouts get logged in a separate fitness tracker. Calories, steps, and progress all exist in completely isolated systems that never talk to each other. Every day starts with the same mental overhead - where did I save that recipe? Did I already hit my protein today? What should I eat after that heavy workout yesterday?
HivRa started as a response to that exact problem. The original idea was simple: bring meal planning into one place. But the more the product evolved, the clearer it became that meals alone weren't the problem. The entire wellness routine was broken - not because any single app was bad, but because nothing was connected. That insight became the foundation of HivRa. Not another tracker. A connected wellness system where every action informs the next decision - and nothing lives in isolation.
HivRa doesn't treat each wellness action as an isolated event. Every meal logged, every workout completed, every streak maintained becomes context for the next recommendation - so the system gets more useful the more you use it.
When you set up HivRa, your calorie targets, protein needs, dietary preferences, and fitness goals form your personal context. Your estimated daily calorie needs are calculated from your age, height, weight, sex, and goal - and everything the app recommends builds on that foundation.
Log a workout and HivRa adjusts. The app reads your steps, active calories, and heart rate directly from Apple HealthKit - so your real-world activity feeds into your meal recommendations automatically, without manual entry. The harder you train, the smarter the suggestions.
Your meal plan auto-generates your grocery list. Saved recipes feed back into future meal plans. Weekly planning keeps your nutrition goals connected across the entire week, not just today. Share your grocery list in real time with anyone - both of you can view and edit it simultaneously.
HivRa tracks your longest continuous streak of daily workout and meal logging. Miss a day and the current streak resets - but your personal best stays. Every time you come back, you are chasing your own record.
HivRa replaces the stack of disconnected apps most people cobble together. Here is what is inside.
AI-powered meal recommendations built around your calorie targets, protein needs, dietary preferences, and recent workout activity. Save full recipes and they re-enter your recommendation pool. Build meal plans and they automatically generate your grocery list. Every nutrition decision stays connected to your goals. Up to 15 personalised AI meal recommendations per day, each one built around your actual context.
Log workouts, exercises, sets, reps, and weight. HivRa remembers your last session weights automatically - no lookup needed. It also reads your steps, active calories burned, and heart rate directly from Apple HealthKit, so your real-world activity feeds into your meal recommendations without extra effort. Rate your session intensity when you finish and HivRa stores a recovery signal - so the app knows which muscles to focus on next.
When you request a recommendation, HivRa builds your context from your fitness goal, remaining daily macros, workout intensity, and meal type preference - then sends only that anonymised context to a Groq-powered Llama model. No personally identifiable information ever leaves the app. What comes back is a meal suggestion with macros, ingredients, and prep instructions matched to what your body actually burned.
Generate your grocery list automatically from your meal plan, or build it manually. Share it in real time with anyone - both of you can view and edit the same list simultaneously, powered by Firestore live listeners. Plan meals up to two weeks ahead across breakfast, lunch, dinner, and snacks. Nutrition is calculated automatically against your daily goal as you plan.
HivRa tracks your longest continuous streak of daily workout and meal logging. Your personal best is always visible. Miss a day and the streak resets - but the record stays. Streak history is stored locally on your device so it is always available, even offline.
Everything in one place. Nutrition summary, workout log, streak record, weekly plan, saved recipes, grocery list, and AI recommendations - all from a single dashboard built for the two minutes you spend checking in each morning.
Four core workflows, from your first tap to your next meal
Every screen in HivRa is built around one principle: every action should take less than thirty seconds. Logging a meal, checking your streak, reviewing an AI recommendation, or updating your shared grocery list should feel effortless - because the best wellness app is the one you actually open every day.
HivRa is a solo-built iOS product - designed, architected, and shipped by one engineer. Every technical decision was made around one goal: build a system reliable enough for daily use, smart enough to adapt, and connected enough to replace a stack of separate tools.
AI in HivRa is not a feature bolted on at the end. It is the recommendation layer the entire nutrition system is built around. When a recommendation is requested, HivRa constructs a prompt dynamically from the user's fitness goal, estimated TDEE, remaining daily macros, current workout intensity, and meal type preference. No personally identifiable information is transmitted to the AI provider. Token-aware prompt construction keeps responses fast and cost-efficient. The model running underneath is Groq's Llama, chosen for inference speed that keeps recommendations feeling immediate.
Built entirely in SwiftUI with structured state management and reusable product flows. HealthKit integration reads step count, active calories, and heart rate directly from the device - keeping the workout-to-nutrition feedback loop automatic without manual entry. Core Data handles local persistence for streak history and planning patterns, keeping that data available offline and off the server.
Firebase Authentication handles secure user identity and session management. Firestore provides real-time data sync across meals, workouts, recipes, meal plans, and grocery lists. The shared grocery list feature is built on Firestore's real-time listeners - multiple users can view and edit the same list simultaneously with no refresh and no delay. All API requests are protected by Firebase Authentication token verification server-side.
A Render-hosted backend sits between the iOS app and the Groq API. It handles token-aware prompt construction, structured response parsing, per-user rate limiting enforced via Firestore transactions, and clean integration boundaries between the client and the AI layer. This separation keeps the iOS app lean and gives the backend room to evolve independently.
Every object in HivRa - meals, workouts, grocery lists, recipes, streak records, user goals - is modeled as part of a connected system. A saved recipe re-enters a meal plan. A logged workout influences a meal recommendation. A session intensity rating generates a recovery signal. A completed daily log contributes to a streak. The data model reflects the product philosophy: nothing in wellness exists in isolation.
When you open a workout, HivRa loads your last session weights automatically. No lookup, no manual entry. That default creates a natural progressive overload pattern - you always start from where you actually were, not from scratch.
HivRa does not sell user data or use it for advertising. HealthKit data is never stored on servers - it is read from the device at the time of use only. AI prompts are constructed from anonymised nutritional context only. Server-side rate limiting is enforced per user via Firestore transactions. All data in transit is encrypted via HTTPS.
HivRa was built through a close collaboration between a software engineer and a physician - combining engineering depth, clinical grounding, and a shared frustration with how fragmented wellness tools actually are.
Full-stack software engineer with 5+ years of experience building enterprise-grade healthcare applications in clinical environments governed by HIPAA compliance standards. That background in healthcare systems, data integrity, and user-centred product thinking directly shaped how HivRa was built - from architecture to the smallest UX decision.
HivRa is her first independently shipped iOS product. The entire architecture - SwiftUI frontend, Firebase backend, Render-hosted API, Groq Llama integration, and HealthKit data pipeline - was designed and built end-to-end as a solo engineering effort.
Other live projects include DocChat, a RAG-based document chat application built with ChromaDB and vector search, and Resume Tailor, an AI-powered resume optimisation tool. Also a contributor to published research on lung cancer detection using machine learning models.
Physician with experience across inpatient and outpatient medicine, preventive health, and patient-centred care.
Dr. Carolina joined HivRa as Clinical Advisor to ensure the product was grounded in how people actually think about their health - not how engineers assume they do. Her input shaped core product decisions throughout development: the workout progression model, the recovery signal concept, meal planning workflows, and the overall UX philosophy.
Every screen was reviewed through her lens - not for clinical accuracy alone, but for whether a real person would find it intuitive, approachable, and genuinely useful in their daily routine. Her clinical perspective is also the reason HivRa never crosses into medical advice territory while still being genuinely helpful for people building healthier habits. That balance was a deliberate product decision, informed by her throughout.