Inspiration
Most of us lose money in the market not because the market is rigged, but because we are human. We sell too early because we are scared and we hold too long because we are hopeful. This invisible conflict creates what we call the Human Tax. We were inspired to build a tool that finally makes this tax visible, transforming the messy and emotional reality of trading into a cold mathematical audit that reveals exactly how much our psychology is costing us.
What it does
Sigma Capital is a premium behavioral analytics suite that audits trade history for thirteen psychological biases. By simply uploading a CSV log, users can see their performance gaps, identify exactly where they are chasing peaks, and understand their true portfolio risk. The platform provides a set of rule based prescriptions that help traders fix their most expensive habits along with a cinematic year end review that tells the story of their 2026 investment journey.
How we built it
The platform is powered by a high performance stack designed for speed and precision. The frontend is built with React and Vite, utilizing a custom dark mode design system and glassmorphism components. For data visualization, we integrated Recharts to handle complex drawdown and allocation histories. The analytical core is a FastAPI backend written in Python, which processes financial time series data using Pandas and NumPy to calculate advanced metrics like the Herfindahl Hirschman Index for concentration and Sharpe ratios for risk adjustment.
Challenges we ran into
One of the most difficult parts of the project was accurately simulating the rational counterfactual universe. We had to ensure that the passive hold baseline was perfectly synchronized with the active trade timestamps to calculate the true behavioral alpha. We also navigated several technical hurdles, from resolving port conflicts in our local development environment to fixing stubborn indentation errors in our Python logic that were silently crashing our risk engine. Ensuring that the peak to trough resilience charts were both accurate and visually readable required multiple rounds of refinement.
Accomplishments that we're proud of
We are particularly proud of the Sigma Coach feature, which successfully translates complex mathematical analysis into human readable advice without relying on external AI APIs. Creating a cinematic Wrapped experience that dynamically summarizes an entire year of trading data in a high fidelity narrative was another major milestone. Seeing the platform transition from a simple data parser into a professional capital management terminal with a unified dark mode aesthetic was incredibly rewarding.
What we learned
Building Sigma Capital taught us that data visualization is about more than just numbers; it is about storytelling. We learned how to manage complex financial models and built a resilient multi tab interface that maintains state across different diagnostic views. More importantly, we discovered the depth of psychological impact on financial outcomes and how to use quantitative indices like HHI to measure concentration bias in a way that feels actionable for a retail investor.
What's next for Sigma Capital
The future of Sigma Capital involves expanding beyond static CSV uploads into direct multi broker API integrations for real time behavioral tracking. We plan to implement more advanced machine learning models that can identify even subtler patterns in trading psychology, such as emotional fatigue or revenge trading. Eventually, we hope to build a community benchmark feature that allows users to see how their behavioral resilience stacks up against the rest of the market anonymous comparison.
Log in or sign up for Devpost to join the conversation.