StackSurge: Investor & Startup Collaboration Platform

StackSurge

Inspiration

Raising capital is one of the biggest challenges for early-stage startups, while investors often struggle to find trustworthy and data-backed opportunities. StackSurge was born from the idea of simplifying this connection by creating a secure, intelligent, and data-driven platform to bridge the gap between innovative startups and interested investors.

What it does

StackSurge is a platform that connects Investors and Startups through tailored dashboards and smart matchmaking. Users can register based on their role and access specific tools:

Startups can pitch ideas, track funding status, and connect with investors.
Investors can explore investment opportunities, monitor ROI, and discover startups based on preferences.
Machine Learning predicts startup profitability and matches users based on goals, industries, and funding needs.
Secure payments via Stripe and document verification add an extra layer of safety.

How we built it

Frontend:

Built with Next.js, TypeScript, Tailwind CSS, and UI libraries like Shad CN UI, Material UI, and Recharts for visualizations.

Backend:

Used Convex for real-time data and database management.
Kinde Authentication for secure role-based access.
UploadThing for file handling.
Stripe for payment integration.

Machine Learning:

Created a Random Forest Classifier using Scikit-learn, trained on isProfitable.csv with financial and growth metrics.
Tuned with GridSearchCV.
Deployed via FastAPI on Render.
ML features include:
- Profitability prediction.
- Investor-startup auto-matching using sector, stage, funding history, and geography.
- Fraud detection using document verification.

Challenges we ran into

Integrating ML with a web-based frontend and backend stack required multiple iteration cycles.
Matching investors and startups in a meaningful way using real data was a non-trivial problem.
Fine-tuning the ML model and ensuring accuracy without overfitting.
Securing user data and implementing role-based authentication in a scalable way.

Accomplishments that we're proud of

Successfully deployed a machine learning model into production using FastAPI.
Created a seamless dual-dashboard experience for two completely different user roles.
Enabled secure and real-time transactions between startups and investors.
Built a fully functional MVP with predictions, user matching, and payment integration.

What we learned

How to architect a scalable and secure full-stack application from scratch.
The importance of designing separate flows for different user types.
The challenges of deploying and tuning ML models in real-world use cases.
How user experience and clear visual data insights improve decision-making for both investors and startups.

What's next for StackSurge

Expanding the ML model to include more complex datasets and features (e.g., sentiment analysis of business plans).
Adding a chat/messaging system for direct investor-startup communication.
Improving fraud detection using advanced NLP techniques.
Launching mobile versions for broader accessibility.
Partnering with accelerators and investment firms to grow the platform user base.

Tech Stack Overview

Frontend: Next.js, TypeScript, Tailwind CSS, Shad CN UI, Material UI, Recharts
Backend: Convex, FastAPI
Authentication: Kinde Auth
File Uploads: UploadThing
Payments: Stripe
Machine Learning: Python, Scikit-Learn, Pandas, NumPy
Deployment: Render (ML API)