Inspiration

"Every decision feels heavy because modern life demands too many of them."

I noticed that people spend hours, even days, paralyzed by choices, often mistaking this indecision for "careful consideration." In reality, every unresolved decision consumes vital mental energy, leading to cognitive resource depletion and lost momentum.

I was specifically inspired by the idea that "doing nothing" is also a decision. Most productivity tools overlook the "Status Quo" option, but inaction has its own hidden costs—missed opportunities and mental fatigue. I built Kairos Flow to help people visualize their choices not as static "Yes/No" questions, but as dynamic timelines, helping them decide not just what to do, but when to do it.


Gemini API Integration

Kairos Flow leverages the Gemini 3 API as its core reasoning engine to transform abstract anxieties into structured, actionable data. The integration is central to the application through these key features:

  • Contextual Future Modeling: Gemini 3 analyzes the user's dilemma and difficulty level to generate three distinct scenarios: Option A and B (viable future paths) and Option C (the "Inaction Path" visualizing the cost of doing nothing).
  • Dynamic Timeline Scaling: I utilized Gemini's reasoning to automatically adjust the temporal scale of the future timeline (Near, Mid, and Long-term) based on the decision's complexity.
  • Actionable Intelligence: The API breaks down complex resolutions into four immediate, manageable To-do items to jumpstart momentum and prevent procrastination.

I used Google AI Studio to rigorously test and refine the system instructions, ensuring the model returns structured JSON. This allowed for seamless data integration within the Flutter frontend, turning open-ended thoughts into a visualizable path for personal growth.


What it does

Kairos Flow is an AI-powered decision support app designed to help users break free from overthinking and transition into decisive action.

Input & Self-Assessment

Upon entering the app, users encounter a text input area to type out their concerns, a process that naturally encourages them to organize and clarify their thoughts. Users then quantify their burden by setting a difficulty level from 1 to 5, allowing them to objectively assess the weight of the dilemma.

Three-Path Future Modeling

Users have the flexibility to either manually write their own options or let Gemini 3 generate them based on the context of their concern. The final selection screen presents three distinct future paths:

  • Option A and Option B — both represent different but equally viable and executable future outcomes
  • Option C — the "Inaction Path" that visualizes the future if the user chooses to do nothing

Dynamic 3-Stage Future Visualization

Once an option is chosen, Kairos Flow reveals a structured timeline consisting of three distinct phases: Near Future, Mid Future, and Long-term Future. To ensure realistic projections, the scale of this timeline is intelligently adjusted based on the complexity and difficulty level of the specific decision.

Execution via Immediate To-Do Lists

To combat post-decision paralysis, the AI generates a concrete 4-step To-do list the moment a final choice is made. These immediate action items are designed to build momentum and prevent users from falling back into a cycle of procrastination even after a decision has been reached.

Personal Growth Archive & Feedback Loop

Users can archive their decisions and To-do lists to create a history of their personal journey. This feature allows future users to recall their past thoughts and mindset during times of struggle. Kairos Flow supports deeper reflection through a commenting system for self-feedback and provides a visual "Completion" button to mark achieved goals and celebrate progress.


How I built it

Frontend: Flutter

Kairos Flow is built with Flutter to provide a seamless and smooth cross-platform mobile experience. I chose Flutter to ensure that users can access decision support anytime, anywhere, especially when they feel the "hidden cost of indecision" in their daily lives.

On the frontend, I utilized Flutter's reactive UI to create a structured, step-by-step flow, including dilemma input, difficulty selection (1–5 scale), dynamic timeline views, and a reflection archive with comment and completion tracking. This architecture allows the interface to update in real-time as AI content arrives, keeping the experience highly interactive.

Intelligence Engine: Gemini 3 API

The core reasoning engine of Kairos Flow is the Gemini 3 API. Rather than using AI as a simple chatbot, I engineered a sophisticated prompt system that transforms open-ended concerns into structured data. To ensure the highest level of accuracy, I utilized Google AI Studio's "Chat with models" feature to rigorously test and refine the system instructions before integrating them with the app's backend.

My system is designed to:

  • Contextual Language Detection — Respond naturally in the user's preferred language
  • Timeline Scaling — Distinguish between emotional severity and real-world duration to set appropriate timeline lengths (Near, Mid, and Long-term)
  • Three-Path Modeling — Simultaneously analyze two viable future paths and one inaction path (Option C)
  • Actionable Output — Generate exactly four concrete, immediate action steps to prevent procrastination

Structured Logic and Reliability

To ensure technical reliability, I implemented strict output formatting rules so Gemini returns clean, structured JSON. This allowed the app to consistently transform abstract "mental fatigue" into visualizable future scenarios. By integrating Gemini 3 as a "Decision-thinking partner," Kairos Flow effectively analyzes context, models long-term consequences, and guides users toward practical next steps.


Challenges I ran into

The Solo-Development Marathon

As a solo developer, I was responsible for the entire product lifecycle—from initial conceptualization and UI/UX design to full-stack development and QA testing. Balancing the creative demands of design with the technical rigors of implementation was a significant challenge. It required strict time management and the ability to switch mindsets constantly between a product manager, a designer, and an engineer.

The Flutter Learning Curve

This project marks my very first experience building a complete app service with Flutter. While Flutter's reactive framework is powerful, mastering its state management and ensuring a smooth, responsive UI for complex features—like the dynamic future timeline—presented a steep learning curve. Overcoming these technical hurdles through rapid self-study and iterative prototyping was one of the most demanding yet rewarding parts of the journey.

Designing for Long-Term Retention

One of the biggest challenges was moving beyond a one-time utility. I spent a significant amount of time deeply considering how to ensure users return to Kairos Flow rather than using it as a one-off tool. This led to the development of the "Reflective Archive" and "Self-Feedback Loop." By allowing users to comment on their past dilemmas and track their progress with the "Completion" button, I aimed to transform the app into a continuous personal growth companion where users can find value in revisiting their decision history and learning from their own evolution.


Accomplishments that I'm proud of

Successful Integration of Complex AI Logic into a First-Time Flutter App

Despite this being my first time developing with Flutter, I successfully integrated the Gemini 3 API to handle sophisticated, multi-turn reasoning. I am proud of creating a stable mobile environment where abstract dilemmas are reliably transformed into structured, visualizable data.

The "Inaction Path" (Option C) Conceptualization

One of my proudest achievements is the conceptualization of Option C—the future scenario for inaction. I successfully turned the abstract "hidden cost of indecision" mentioned in my vision into a tangible UI element, helping users realize the opportunity costs of not deciding.

Designing a Meaningful Feedback Loop for Personal Growth

Beyond just solving a problem, I built a system that encourages long-term self-reflection. I take pride in the "Reflective Archive" feature, which allows users to leave comments on their past selves and experience the satisfaction of hitting the "Completion" button as they grow.

Optimization of Mental Battery Life

I am proud to have built a tool that actively preserves the user's "mental battery." By delegating the heavy lifting of scenario-modeling to Gemini 3, I successfully created a "Decision-thinking partner" that helps users save their cognitive resources for actual execution.


What I learned

The Power of AI as a Decision Partner

Through this project, I gained deep insights into the capabilities of the Gemini 3 API. I learned how to move beyond simple "question and answer" interactions by implementing structured prompt engineering that allows the AI to act as a "reasoning engine" for complex human dilemmas. I discovered that Gemini 3 is exceptionally capable of modeling future scenarios and breaking down abstract worries into concrete, actionable data.

Developing with Empathy

This was my first experience developing a product specifically aimed at reducing human stress and enhancing mental well-being. Throughout the process, I kept the future user in mind, focusing on how Kairos Flow could alleviate their mental fatigue and "decision paralysis." I learned that the most rewarding part of development is not just the code itself, but the feeling of knowing my work can genuinely help someone regain their momentum and peace of mind.

The Intuitive Efficiency of Flutter

As a first-time Flutter developer, I was impressed by its intuitive structure and efficiency. I learned that Flutter's reactive framework is ideal for building fast, beautiful, and accessible apps that people can use anywhere to manage their mental energy. This project has given me the confidence to continue building cross-platform solutions that solve real-world problems.


What's next for Kairos Flow

Official App Store Deployment

While Kairos Flow is currently deployed as a web application using Firebase, my immediate next step is to officially release it on both the Apple App Store and Google Play Store. This will ensure that users can have more stable and native access to decision support directly on their mobile devices.

Data-Driven User Onboarding

I plan to transition from local-only storage (Hive) to a cloud-based database system. By analyzing anonymized data on common dilemmas, I will introduce a "Trending Concerns" section on the home screen. This will allow users to jumpstart their decision-making process with "Quick Start" templates for frequently shared worries, significantly increasing app engagement and daily utility.

Multimodal Input: Voice and Vision

To make the app even more accessible, I will integrate Voice Recognition (STT) and Image Upload capabilities. Sometimes, dilemmas are too complex to type out or require visual context. These features will allow Gemini 3 to understand nuanced situations that are difficult to describe with text alone, providing a more empathetic and comprehensive analysis.

Advanced Prompt Engineering

I will continuously refine and strengthen our AI prompt architecture. My goal is to leverage the full potential of Gemini 3 to provide even more granular, detail-oriented, and personalized future scenarios, ensuring that every user receives advice that feels specifically tailored to their unique life context.

Built With

Share this project:

Updates