Inspiration Firsthand experience with time-consuming quantitative market research and risk management in professional finance inspired us to build a more efficient solution.
What it does
Portfolio Wizard orchestrates multiple AI agents to automate financial research, perform quantitative analysis such as backtest trading strategies, and provide real-time risk management through a unified platform.
How we built it
We built a Next.js application deployed on Vercel, integrated Clerk for authentication, leveraged OpenAI's whisper for speech to text, Llama 7B hosted on Groq for inference, used Exa to equip our AI agents with web search capabilities, and implemented ElevenLabs for text to speech.
Challenges we ran into
Streaming complex financial data and orchestrating multiple AI agents to perform intricate analysis & complex calculations while maintaining low latency in real-time
Accomplishments that we're proud of
Orchestrated and engineered a multi-agent system that reduces research time from hours to minutes, and which is capable of quantitative analysis and risk management.
What we learned
Learned how to architect and orchestrate an agent gateway, classifier and coordinate multiple AI agents to perform complex financial research and quantitative analysis.
What's next for Portfolio Wizard
We will equip Portfolio Wizard with a wide range of tools to perform additional quantitative functions and we will allow users to input their risk tolerance, add a watch list of stocks, send alerts and create a usage based model to generate revenue.
Log in or sign up for Devpost to join the conversation.