Smarter investment insights
Powered by a team of AI agents
⚡Orchestrated by you ⚡
Retrieval Augmented Generation (RAG) provides means for feeding external data to the AI agents for processing.
When multiple AI nodes (or agents) collaborate on a task, they can achieve significantly greater outcomes—a concept that has been substantiated through various research efforts.
The architecture employs a directed acyclic graph (DAG) structure, making it inherently linear. This approach distinguishes it from other agent-driven solutions in the industry. As the orchestrator, you determine which agents handle specific tasks and the sequence in which they are executed.
Each agent is given a very focused mandate (instruction).
Agents can make API calls to Yahoo Finance to enhance the timeliness and of their evaluations.
AVA leverages the llmware library, enabling it to run local LLMs like LLaMA and cloud-based LLMs like ChatGPT and Claude.
Easily share your insights. Allow others to read your conversation and outputs from the AI team.
.jpg)
Answers to questions you might have about AVA.
AVA specializes in active investment advice focused exclusively on single stocks. The system leverages advanced financial metrics like the Piotroski F-Score to evaluate and recommend stocks based on real-time data and user-defined parameters. It does not offer portfolio management, mutual fund advice, or passive investment strategies.
AVA employs an agentic Retrieval-Augmented Generation (RAG) architecture. This integrates real-time data retrieval with Large Language Models (LLMs) to ensure that recommendations are both contextually accurate and up-to-date. The system uses specialized agents to process financial data, evaluate risk profiles, and deliver personalized advice.
As of December 2024, AVA primarily utilizes the Piotroski F-Score, a proven financial scoring method that evaluates companies based on profitability, leverage, and operational efficiency. This metric helps identify undervalued stocks with strong financial positions, but the architecture is flexible and can incorporate alternative metrics in the future.
AVA is designed for investors seeking active advice on individual stocks and those comfortable making decisions based on detailed financial analysis. It is ideal for users looking to independently manage their stock investments while receiving informed recommendations. It may not be suitable for investors seeking comprehensive financial planning or risk-averse strategies like index funds.
AVA prioritizes transparency and fairness by clearly explaining its recommendations and ensuring alignment with user-defined risk profiles. It incorporates ethical safeguards to prevent biases in its algorithms and continuously updates its assessments to reflect changing user goals and market conditions.
