Picking Geek AI — Investing Made Easy with NVIDIA-Powered AI.

Inspiration

Professional investment research tools are often expensive and complicated. We built Picking Geek AI to make AI-powered stock analysis accessible to everyone, helping individual investors make smarter, more informed decisions.

What it does

Picking Geek AI is an AI-powered stock research platform supporting U.S. and Taiwan stocks.

My Stock: Track stocks using MA, MACD, Bollinger Bands, and Squeeze indicators, with customizable alerts.
AI Invest Pro: Fast AI analysis powered by NVIDIA Nemotron-3-Nano-30B-A3B.
AI Invest Ultra: Advanced analysis powered by NVIDIA Nemotron-3-Ultra-550B-A55B, including competitor comparisons and relevant news.
Search & Personalization: Stock search, multilingual support, and appearance settings.

How we built it

Built with a Python backend, financial market APIs, and deployed on Render. We integrated NVIDIA Nemotron models through Nebius, combining financial data with retrieval-augmented AI analysis (RAG) to generate structured investment insights.

Challenges we ran into

Our biggest challenges were integrating financial data with AI reasoning, balancing inference speed and analysis depth, and presenting complex technical indicators in a user-friendly interface.

Accomplishments that we're proud of

We built and deployed a working AI-powered investment platform, integrating two NVIDIA models into a unified research experience. As an independent developer, turning this vision into a functional product is our proudest achievement.

What we learned

We learned that effective financial AI requires reliable data, efficient model selection, and clear explanations. Powerful AI should enhance human investment decisions, not replace them.

What's next for PickingGeek

We plan to introduce multi-agent financial research, faster market data processing, enhanced RAG capabilities, and personalized investment insights. Our mission: Investing Made Easy with AI

Built With

django
nebius
neon
postgresql
python
render

Submitted to

Nebius x NVIDIA Global AI Hackathon

Created by

As the solo founder and developer of Picking Geek AI, I led the entire project from concept to deployment.

My contributions included product ideation, system architecture, full-stack development, Python backend implementation, financial API integration, NVIDIA Nemotron model integration through Nebius, prompt engineering, UI/UX design, testing, and cloud deployment.

I also managed the product branding, documentation, and hackathon demo production, leveraging AI tools such as ChatGPT, Codex, and Gemini to accelerate development.

From idea to working product, I independently built and delivered Picking Geek AI — Investing Made Easy.

James Chen

Updates

James Chen started this project — Oct 19, 2026 12:59 AM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.