MeshFlow

Inspiration

We were inspired by Bloomberg Terminals. Professional investors rely on these to analyze markets, but they're expensive and complicated to use, making them inaccessible to the average person, and the alternatives are also either expensive or only focus on specific parts of the market. So, we thought of MeshWise, an AI Agent that would make investing easy and accessible.

What it does

MeshWise uses many AI-agents who all do research on different parts of the market, from historical analysis to scanning modern news stories for market sentiment. Then, one final agent consolidates all of these findings to give the user a simple, all-encompassing answer, giving the opportunity to anyone with any level of financial literacy a gateway to investing.

How we built it

It was built using the Solace Agent Mesh, where we made custom AI agents using Yaml and Python. We also utilized the YellowCake API for the modern news research agent.

Challenges we ran into

There was a definite learning curve in understanding how to use the Solace suite of products. The biggest challenges were getting the SAM gateway to work as well as getting the YellowCake agent to work inside of Solace and produce good results

Accomplishments that we're proud of

We're proud of making several AI agents that all have their place in the project, all together making one smooth flow for the experience. We're also proud of getting the SAM gateway to work as well as getting the YellowCake research agent to be able to do research from within the Solace flow.

What we learned

We learned a lot about prompt engineering and how to get the agents to do specific jobs. We learned about distributed systems through the Solace Agent Mesh.

What's next for MeshFlow

We want to continue to improve the investing flow for the user by fine-tuning the current agents we have to consistently return good answers at a good pace. We want to add more agents as well as add machine learning models into the decision-making process. We want to add more features to the UI, such as a display of the user's portfolio, as well as add a database to store user information to allow for a more personal experience.