Aaron AI (RESOLV.AI) 🚀

The Agentic Financial Bodyguard & Goal Guardian

To check the Opik traces after running the AI agent

Go to https://www.comet.com/opik/api and login with your credentials

Aaron AI (internally branded as RESOLV.AI) is an autonomous financial coaching system built for the 2026 New Year Resolution Hackathon. Unlike passive budgeting apps, Aaron AI uses a multi-agent orchestration to proactively intervene in spending habits, evaluate financial feasibility, and guide users toward their long-term goals using high-fidelity AI observability.

Here is one of our traces id

019c25c3-d5f4-7d4a-9953-089c87868dcc

---

🌟 Key Features

• Proactive Goal Tracking: Moves beyond spreadsheets to provide active "nudges" based on user-defined financial missions.
• Multi-Agent Orchestration: Powered by the Google Agent Development Kit (ADK) to separate reasoning, math, and coaching logic.
• Production-Grade Observability: Deep integration with Opik for real-time trace logging, response evaluation, and LLM-as-a-judge scoring.
• Behavioral Intervention: Custom logic that factors in risk appetite and income predictability to provide personalized advice.

---

🛠️ Tech Stack

• Frontend: Next.js 14+, Tailwind CSS, Lucide React.
• Backend: FastAPI (Python 3.13), Uvicorn.
• AI Engine: Google ADK, LiteLLM.
• LLMs: Llama 3.3 70B & 3.1 8B (via Groq).
• Observability: Opik (by Comet).

---

🚀 Getting Started

1. Prerequisites

Ensure you have the following installed:

• Node.js (v18+)
• Python (v3.10+)
• Git

--- AI & LLM CONFIGURATION ---

Get your key at https://console.groq.com/

GROQ_API_KEY=your_groq_api_key_here

--- OPIK OBSERVABILITY ---

Get your key at https://www.comet.com/docs/opik/

OPIK_API_KEY=your_opik_api_key_here OPIK_PROJECT_NAME=Aaron OPIK_WORKSPACE=your_workspace_name

GEMINI_API_KEY="your_gemini_api_key_here"

# Opik Configuration
OPIK_API_KEY="your_opik_api_key_here"
OPIK_URL_OVERRIDE="https://www.comet.com/opik/api"
OPIK_PROJECT_NAME="Aaron"
OPIK_WORKSPACE="your_opik_workspace_here"

# LLM Provider API Keys (if needed)
OPENAI_API_KEY="your_openai_api_key_here"
OPENAI_API_KEY_2="your_openai_api_key_2_here"
GROQ_API_KEY="your_groq_api_key_here"

# Logging Configuration (see Debug Mode and Logging section below)
OPIK_CONSOLE_LOGGING_LEVEL="WARNING"  # Python: Control console output (DEBUG, INFO, WARNING, ERROR, CRITICAL)
OPIK_FILE_LOGGING_LEVEL="DEBUG"       # Python: Enable file logging
OPIK_LOG_LEVEL="DEBUG"                # TypeScript: Control log level

--- SERVER CONFIGURATION ---

PORT=8000 FRONTEND_URL=http://localhost:3000

Environment Configuration 🔑

Before running the backend, you must set up your environment variables.

1. In the root directory, copy the example file:

   cp .env.example .env
   

2. Installation & Frontend Setup

Clone the repository and install dependencies for the user interface.

Note: If you encounter any bugs during the frontend setup, please create a new issue on GitHub.

# Clone the project
git clone [https://github.com/Nailer/aaron-AI](https://github.com/Nailer/aaron-AI)

# Navigate to the project directory
cd aaron-AI

# Install dependencies
# We use NPM for package management. See the official [NPM Documentation](https://docs.npmjs.com/) for more info.
npm install

# Start the frontend development server
npm run dev

3. AI Backend & Opik Observability Setup

To enable the AI "brain" and log traces for response evaluation, follow these steps to set up the Python environment.

# Initialize Virtual Environment
# Create and activate a virtual environment to keep dependencies isolated.

# For macOS/Linux:
python3 -m venv venv
source venv/bin/activate

# For Windows
python -m venv venv
.\venv\Scripts\activate

Install Requirements & Start API

# Install backend dependencies
pip install -r requirements.txt

# Start the FastAPI server
uvicorn backend.api:app --port 8000 --reload

Enable Agent Tracking

In a new terminal tab (with the venv activated), run the tracking script to initiate the Opik logging session:

python backend/agent_tracking.py

# All agent trajectories, tool calls, and LLM outputs will now be logged to the Opik dashboard for evaluation.

📊 AI Observability with Opik

We utilize Opik to ensure our financial advice is accurate and safe.

• Tracing: Every interaction is traced through the Google ADK runner.

• Evaluation: We use LLM-as-a-judge metrics to score "Financial Responsibility".

• Debugging: Traces allow us to deduce exactly why an agent recommended a specific savings plan.

├── backend/ │ ├── api.py # FastAPI Endpoints │ ├── agent_runner.py # ADK Runner logic │ ├── agent_tracking.py # Opik integration script │ └── init.py # Financial calculation tools ├── frontend/ # Next.js Application ├── requirements.txt # Python dependencies └── package.json # Node.js dependencies

🛡️ Security & Privacy

Aaron AI is built with financial privacy in mind. We do not store raw bank credentials. All data processed is used strictly for the generation of your financial plan and is encrypted during transit.

📄 License

Distributed under the MIT License. See LICENSE for more information.

Developed with ❤️ for the 2026 Comet Hackathon.