Voice Assistant And Multimodal Demo

This repository contains two AI applications:

Project	Folder	Description
Vault-AI	Vault-AI/	A banking MCP server built with FastMCP, paired with a LangGraph ReAct agent and Streamlit UI.
Voice-AI-Bot	Voice-AI-Bot/	A multimodal Streamlit chatbot powered by Google Gemini with Speechmatics STT/TTS support.

Both projects share a single virtual environment and a common requirements.txt.

---

1. Prerequisites

• Python 3.10+ (3.12 recommended)
• ffmpeg installed and available on your PATH (required for Voice-AI-Bot)

Install ffmpeg

macOS

brew install ffmpeg

Linux

sudo apt install ffmpeg

Windows

Download from: https://ffmpeg.org/download.html

Verify installation:

ffmpeg -version

---

2. API Keys

You'll need the following API keys (all offer free tiers):

• Groq — Used by the LangGraph agent LLM
  https://console.groq.com/keys

• Google AI Studio (Gemini)
  https://aistudio.google.com/app/apikey

• Speechmatics — Speech-to-Text and Text-to-Speech
  https://portal.speechmatics.com/

---

3. Setup

From the project root, run:

./setup.sh

If you have uv installed, you can use:

./setup.sh uv

This creates the .venv/ virtual environment and installs all dependencies from requirements.txt.

Next, create your environment file:

cp .env.example .env

Open .env and add your API keys.

---

4. Activate the Environment

In every new terminal session, activate the virtual environment first:

source .venv/bin/activate

---

5. Running the Applications

Vault-AI

Terminal 1 — Start the MCP Server

cd Vault-AI/MCP-Server
python banking_mcp_lab.py

Terminal 2 — Start the Streamlit Agent UI

cd Vault-AI/Agent-Client
streamlit run app.py

---

Voice-AI-Bot

cd Voice-AI-Bot
streamlit run voice_bot_lab.py

---

6. Development Workflow

Each project includes:

• snippets.txt — Code snippets to paste into scaffolded files
• run_commands.txt — Quick reference for startup commands

Follow the === STEP N === markers inside the scaffold files when adding snippets.

---

7. Project Structure

Project/
├── README.md
├── requirements.txt
├── setup.sh
├── .env.example
│
├── Vault-AI/
│   ├── MCP-Server/
│   │   ├── banking_mcp_lab.py
│   │   ├── snippets.txt
│   │   ├── bank_data.db
│   │   └── run_commands.txt
│   │
│   └── Agent-Client/
│       ├── app.py
│       ├── agent.py
│       ├── config.py
│       ├── multimodal.py
│       ├── files/
│       └── run_commands.txt
│
└── Voice-AI-Bot/
    ├── voice_bot_lab.py
    ├── snippets.txt
    └── run_commands.txt

---

8. Troubleshooting

.env file not found

You have not created the .env file yet.

Run:

cp .env.example .env

Then add your API keys.

---

GROQ_API_KEY is missing

Your .env file exists, but the key is missing or invalid.

---

Port 5173 already in use

Another MCP server is already running.

Find and stop it:

lsof -i :5173

---

ffmpeg: command not found

Install ffmpeg using the instructions in the prerequisites section.

---

401 Unauthorized from Speechmatics

Your SPEECHMATICS_API_KEY is invalid, expired, or missing.

---

Happy building 🚀