Rondavu

We created a patent-pending AI decision agent–powered collaboration platform that lowers group decision fatigue and empowers healthier and more inclusive dining experiences by incorporating individual preferences, health signals, and explainable recommendations that all flow in harmony with group messaging. I have spearheaded end-to-end UX strategy, research, and design around this project to develop a human-centric decision system that accounts for dietary goals on an individual level, social dynamics, and real-world constraints.

Drawing on behavioral research, decision modeling, and prototype validation, I designed an experience where an AI decision agent helps mediate group choices while still accounting for individual dietary needs.

Tools Used: Sketching, Maze, Figma, FigJam, Canva, ChatGPT, Replit

What is the problem?

Market research helped me understand the direction the food industry is moving — whether it be changing consumer tastes or meal habits to rising health concerns. It exposed real tensions that people face when deciding where and what to eat, shaped by culture, cost, and regional norms. Those revelations demonstrated disparities between the present design of menus and service models and how people’s habits are evolving, and enabled me to position it as about adapting the food experience to more health-oriented and values-oriented people’s habits today.

What are the goals?

First step was to understand the premise of the product's vision. It started off by one of Rondavu's founders having a whiteboarding session when planning lunch with colleagues, and an idea was born. Our team got together to brainstorm the goals of the project and measures of success of the project.

Goal

Motivate people to keep track of their diet and healthy choices when going out to eat.

Measure

Test features that integrate health applications with Rondavu to help users track their diet progress and lifestyle when going out to eat.

Goal

Make decisions easier when going out to eat with a group of friends, colleagues, or family while tracking one's healthy habits when out. Providing search engines for groups with different preferences and lifestyles to provide the most compatible options.

Measure

Do groups feel like their decision-making has been made easier when going out to eat? Do they get the right recommendations based on what is up and coming and what fits their needs? Does each group member feel like they've been heard? Does each member feel like they are offered the best experience for their preferences, but also the preferences of the whole group? In order to measure this, we participated in a series of group outings and interviewed groups of people on how they manage decision-making when they go out to eat with friends, family, or colleagues.

What are the common lifestyle habits?

Following goal analysis, we mapped behavioral patterns around group eating to understand how health goals intersect with social dynamics and real-world constraints. This synthesis revealed where users abandon healthy intentions—not due to lack of motivation, but due to social friction—guiding us toward designs that support behavior change without guilt or rigidity.

How do people make decisions?

I crafted a decision mind map about how Rondavu produces recommendations that rendered the complexity of group dining visible. The framework distilled varied inputs — dietary needs, health goals, budget, cultural authenticity, sustainability, social dynamics, and logistics — to create a structured decision model. By separating inputs, resolution logic, and outputs, it revealed how individual preferences, group values, and real-world constraints coalesce into a single flow. This strategy identified critical edge cases early and guided when deterministic rules would best help versus AI-assisted reasoning, facilitating explainable recommendations which reduce decision fatigue while providing a clear picture for group priorities.

Who is the target audience?

I explored multiple user personas to understand how group dynamics impact decision-making when eating out. From the health-conscious friend to the budget-restricted diner to the indecisive planner, each persona brought unique challenges — especially when navigating diverse preferences, dietary needs, and time constraints. By mapping their pain points, I identified common friction zones and designed solutions that ease coordination and reduce the mental load of group dining.

What are the user scenarios and possible solutions?

Next we led a collaborative How Might We workshop to translate insights from user research into actionable design opportunities before defining specific scenarios and solutions. This process centered on putting complex behavioral, social, and logistical challenges into clear problem statements the team could come together around. We explored four fundamental facets of the group dining experience—restaurant selection, personalized experience among groups, comparison of choices, and social experience—to figure out where friction, negotiation, and decision fatigue often occur. This activity aided the team in transitioning from isolated pain points to broader system awareness and ensured that following concepts addressed root causes instead of symptoms. The outputs of this workshop informed the scenarios, feature exploration, and solution concepts that followed.

To anchor Rondavu in real-world decision-making, we collaboratively mapped high-frequency user stories that capture the most common group dining scenarios. This exercise helped us define the product context, uncover core pain points, and ideate solution patterns that support healthier choices without disrupting social dynamics.

Keeping Up With Diet

For users with health or weight-related goals, eating out is less about food selection and more about navigating competing pressures—personal intentions, social expectations, and decision fatigue. To unpack this complexity, we identified the core behavioral, cognitive, social, and psychological mechanisms that shape how people make group dining decisions.

Going Out To Eat With Friends

Going out to eat can be tough when scheduling a get-together with a group of people who have their preferences. Rondavu's AI functionality can recommend options within the messaging tool based on what is available according to everyone's preferred settings on the app.

Rondavu embeds AI-powered recommendations directly into the group’s messaging flow, allowing everyone to participate in the decision without leaving the conversation. By factoring in individual preferences, dietary needs, and real-time availability, the system surfaces options that are most likely to reach group consensus—fast.

Value Proposition

This value proposition workshop synthesized user research, behavioral insights, and business considerations into a clear, shared understanding of Rondavu’s purpose and differentiation. By mapping user pain points—such as decision fatigue, social pressure, and conflicting preferences—against user and business value, the exercise clarified how Rondavu’s conversational AI mediates group consensus rather than simply generating recommendations. The outcome aligned stakeholders around a group-first decision model that reduces friction, supports healthier and more inclusive choices, and naturally fits into existing social conversations. This framework not only articulated a compelling user promise but also established a scalable foundation for growth, partnerships, and future feature prioritization.

How do we differentiate from competitors?

The goal of the competitive analysis was to understand how existing platforms support (or fail to support) group-based decision-making around dining and social planning. We evaluated leading apps to identify where they excel in individual discovery—and where they break down when preferences, constraints, and coordination shift from individual to group contexts.

We analyzed each platform’s core features, strengths, and limitations through the lens of group coordination, speed to consensus, and contextual relevance. This allowed us to identify clear opportunity areas where Rondavu could differentiate by supporting shared decision-making rather than isolated search.

What is MVP?

After reviewing competitive gaps and feedback on existing solutions, the team facilitated an affinity-based prioritization workshop to align on feature scope. Features were evaluated and grouped based on user impact, urgency, and technical feasibility, resulting in a clear MVP cutoff and a shared roadmap for POC exploration.

Let's brainstorm!

Sketching and ideation played a crucial role in visualizing the core features of the app early in the design process. These quick, low-fidelity sketches helped translate abstract ideas into tangible layouts, making it easier to identify key user interactions and refine the app's structure. By iterating on paper, I could explore design options that gave me confidence when traslating to hi-fi mockups. The bigest piece was the messaging application and how voting was constructed along with AI functionality to list the options, and information about each place in review. It is still being refined and I keep coming back to pencil and paper throughout the process.

Vibe Coding

After initial sketching, we used rapid prototyping in Replit to explore and pressure-test interaction ideas through live AI-enabled chat flows. This approach helped us quickly evaluate feasibility, refine behaviors, and uncover new opportunities beyond static designs.

What is the user journey and experience flow?

After the user journey mapping, a messaging functionality and in-app experiences displayed in a user flow diagram was helpful in understanding the flow of the experience when users want to take actions. They start with the messaging app and talk about going out to eat. Rondavu icon in the tray menu can bring up AI search tool to suggest the best places to go to in accordance to everyone's preferences.They can view information, menu options, and discounts before ordering Uber to pick them up. They also get to vote on which option they prefer.

• Established a clear UX strategy and high-fidelity prototype that articulated the product’s value proposition, aligned cross-functional stakeholders, and served as a shared blueprint for subsequent development phases.

• Gathered early user validation confirming the experience felt intuitive and engaging, increasing confidence in the product direction.

• Facilitated cross-functional collaboration to align product, brand, and technical considerations into a cohesive roadmap.

• Designed investor-facing UX assets—including a product walkthrough video, pitch deck, and one-pager—that translated the product experience, value proposition, and roadmap into clear, decision-ready narratives. These materials aligned founders and stakeholders, reduced ambiguity in investor conversations, and supported ongoing fundraising and partnership efforts.
  
  
  

“It is a lot like dining out with a bunch of friends.

Where to go? Italian? Vegetarian? Steakhouse?

Each person has their own appetite, which can make creating a 'shared- interest' extremely challenging.”

"I would like to have the option to select invitees from a friends list, a list of possible restaurant options, available dates and times for a get-together, and send out. Friends would input their choices and rank them……the output would be a list of friends that would attend, a restaurant, time, and date."

• Group dining decisions are driven as much by social pressure as by food preferences.

• Health-conscious behavior has become more intentional and socially influenced post-COVID.

• Users want fewer decisions, not fewer options—simplicity must coexist with personalization.

• Scheduling and location often outweigh cuisine in group settings.

• AI recommendations reduce social friction by acting as a neutral decision mediator.

• Persistent preference tracking enables trust, efficiency, and healthier group outcomes.