Subtle Sense

SubtleSense: AI-Powered Emotion Intelligence

SubtleSense is an AI-powered emotion intelligence platform designed to help people identify and understand emotions that are difficult to recognize, articulate, or express. Instead of focusing only on obvious emotional states, SubtleSense explores subtle emotional patterns and uses AI-assisted analysis to transform personal reflections into meaningful insights.

The Problem

Human communication is rarely emotionally straightforward. People often experience mixed emotions, emotional suppression, uncertainty, indirect expression, and rapidly changing moods that cannot be captured through simple labels such as “happy” or “sad.” SubtleSense explores how artificial intelligence can help users recognize these less obvious emotional patterns while keeping the experience understandable and accessible.

The Product

The platform combines AI-powered journaling, emotion analysis, mood visualization, interactive mental-wellness tools, and personalized insights within a single application. Users can record thoughts and experiences through natural language, explore emotional patterns over time, visualize their mood, and reflect on changes in their emotional state.

SubtleSense is designed as a self-reflection tool, not a replacement for professional mental-health support. Its purpose is to provide users with another way to organize their thoughts and notice patterns they may otherwise overlook.

Building SubtleSense

I built SubtleSense independently from the ground up, including its interface, core functionality, database architecture, AI integration, and overall product experience. The application was developed using React, TypeScript, Tailwind CSS, Superbase, and Google Gemini, with the system designed around interactive user experiences and AI-generated insights.

Rather than stopping at a technical prototype, I developed SubtleSense into a publicly accessible product and continuously iterated on it through experimentation, feedback, and real-world usage.

Traction & Iteration

SubtleSense has attracted 700+ organic visitors and has been launched and shared through platforms including Product Hunt and developer communities. The project has evolved through multiple iterations, with new features, improvements to the user experience, and changes driven by how people actually interact with the product.

Building for real users introduced a different challenge from simply making an application work. It required thinking about usability, retention, product positioning, accessibility, reliability, and the gap between what an AI system can technically predict and what users actually find useful.

Research & Responsible AI

SubtleSense also became the foundation for my broader interest in affective computing and responsible emotion AI. I became particularly interested in the limitations and biases involved in systems attempting to interpret human emotion, including the influence of cultural context, language, data quality, ambiguous expression, and model assumptions.

This led me to investigate emotion-recognition systems beyond simply building an application. I wanted to understand not only whether AI could recognize emotion, but when and why it could fail.

Why I Built It

SubtleSense represents my attempt to combine software engineering, artificial intelligence, human-centered design, product development, and independent research into one continuously evolving project.

More importantly, it reflects how I approach problems: identify a difficult real-world problem, build a working system from scratch, expose it to real users, study its limitations, and iterate based on evidence. SubtleSense is therefore not just an application I built. It is an ongoing experiment in understanding how technology can interact with one of the most complex things humans produce: emotion.