AI Conversational Agents Portfolio

A collection of two production-ready AI-powered conversational agents built with Python FastAPI and Google Gemini AI. These projects demonstrate full-stack development, AI integration, and modern web technologies.

Projects

1. Company Research Assistant

An AI agent that researches companies and generates comprehensive account plans.

Features:

AI-powered company research using Google Gemini
8-section account plan generation
Interactive chat for follow-up questions
Real-time progress tracking
Voice and chat modes

Use Cases:

Sales teams researching prospects
Business development planning
Market intelligence gathering
Competitive analysis

View Project →

2. Interview Practice Partner

An AI agent that conducts mock interviews and provides detailed feedback.

Features:

6 different interview roles (Software Engineer, Sales, Product Manager, Data Scientist, Retail, Marketing)
Dynamic, adaptive questioning
Comprehensive performance feedback
Interview transcripts
Voice and chat modes

Use Cases:

Interview preparation
Professional development
Skills assessment
Communication practice

View Project →

Tech Stack

Backend

Python 3.8+ - Programming language
FastAPI - Modern async web framework
Google Gemini AI - AI language model (gemini-2.0-flash-lite-001)
Uvicorn - ASGI server
Pydantic - Data validation
python-dotenv - Environment management

Frontend

HTML5/CSS3 - Structure and styling
Vanilla JavaScript - No frameworks, pure JS
Web Speech API - Voice recognition and synthesis
Fetch API - HTTP communication

Architecture Pattern

Three-Layer Architecture: Frontend → Backend API → AI Service
RESTful API Design
Session-based State Management
Async/Await Patterns

Quick Start

Prerequisites

Python 3.8 or higher
Google Gemini API key (Get free key)
Modern web browser (Chrome/Edge for full voice support)

Setup Both Projects

# Clone the repository
git clone https://github.com/yourusername/ai-conversational-agents.git
cd ai-conversational-agents

# Setup Company Research Assistant
cd company-research-assistant/backend
pip install -r requirements.txt
echo "GOOGLE_API_KEY=your_key_here" > .env
python serve_frontend.py &

# Setup Interview Practice Partner
cd ../../interview-practice-partner/backend
pip install -r requirements.txt
echo "GOOGLE_API_KEY=your_key_here" > .env
python serve_frontend.py &

# Access the applications
# Company Research: http://localhost:8000
# Interview Practice: http://localhost:8001

Project Structure

ai-conversational-agents/
├── company-research-assistant/
│   ├── backend/
│   │   ├── main.py                  # FastAPI application
│   │   ├── serve_frontend.py        # Server with static files
│   │   ├── requirements.txt         # Dependencies
│   │   └── .env                     # API key (not in repo)
│   ├── frontend/
│   │   ├── templates/
│   │   │   └── index.html          # UI
│   │   └── static/
│   │       └── app.js              # JavaScript
│   ├── .gitignore
│   └── README.md
│
├── interview-practice-partner/
│   ├── backend/
│   │   ├── main.py                  # FastAPI application
│   │   ├── serve_frontend.py        # Server with static files
│   │   ├── requirements.txt         # Dependencies
│   │   └── .env                     # API key (not in repo)
│   ├── frontend/
│   │   ├── templates/
│   │   │   └── index.html          # UI
│   │   └── static/
│   │       └── app.js              # JavaScript
│   ├── .gitignore
│   └── README.md
│
├── HOW_AGENTS_WORK.md              # Technical documentation
└── README.md                       # This file

Key Features

Both Projects Include:

✅ Dual Mode Support

Chat Mode: Text-based conversation
Voice Mode: Speech-to-text and text-to-speech

✅ Modern UI

Responsive design
Clean, professional interface
Real-time updates
Visual feedback

✅ AI Integration

Google Gemini 2.0 Flash Lite
Context-aware conversations
Intelligent response generation
Natural language understanding

✅ Production-Ready

Error handling
Input validation
Session management
API documentation
Browser compatibility

Technical Highlights

Backend Architecture

Async Endpoints: Non-blocking I/O for better performance
Type Safety: Pydantic models for request/response validation
Auto Documentation: OpenAPI/Swagger docs at /docs
CORS Enabled: Cross-origin resource sharing configured
Session Management: In-memory storage for demo purposes

Frontend Architecture

Vanilla JavaScript: No framework dependencies
Event-Driven: Responsive UI updates
Web APIs: Native browser features for voice
Modular Code: Clear separation of concerns
Progressive Enhancement: Works without voice in all browsers

AI Integration

Prompt Engineering: Carefully crafted prompts for quality responses
Context Management: Conversation history maintained
Error Handling: Graceful degradation on API issues
Rate Limiting: Prevents API abuse

API Documentation

Both projects include interactive API documentation:

Company Research Assistant:

Swagger UI: http://localhost:8000/docs
ReDoc: http://localhost:8000/redoc

Interview Practice Partner:

Swagger UI: http://localhost:8001/docs
ReDoc: http://localhost:8001/redoc

Browser Compatibility

Full Support (Chat + Voice):

✅ Google Chrome (Recommended)
✅ Microsoft Edge
✅ Opera

Chat Only:

⚠️ Safari (Voice features limited)
⚠️ Firefox (Voice requires manual config)

Environment Variables

Each project requires a .env file in its backend/ directory:

GOOGLE_API_KEY=your_google_gemini_api_key_here

Get your free API key at Google AI Studio.

Documentation

HOW_AGENTS_WORK.md - Detailed technical documentation explaining architecture, data flow, and implementation details
Company Research README - Specific docs for the research assistant
Interview Practice README - Specific docs for the interview partner

Development

Running in Development Mode

# Company Research Assistant
cd company-research-assistant/backend
uvicorn main:app --reload --port 8000

# Interview Practice Partner
cd interview-practice-partner/backend
uvicorn main:app --reload --port 8001

Testing

Both projects can be tested through:

Interactive UI (open HTML files)
API documentation (/docs endpoints)
Direct API calls (curl, Postman, etc.)

Skills Demonstrated

This portfolio showcases:

✅ Full-Stack Development: Frontend, backend, and AI integration
✅ Modern Python: Async programming, type hints, FastAPI
✅ JavaScript: ES6+, async/await, Web APIs
✅ AI Integration: Prompt engineering, context management
✅ API Design: RESTful principles, documentation
✅ UX Design: Responsive interfaces, error handling
✅ Voice Technology: Web Speech API implementation
✅ Best Practices: Code organization, security, performance

Use Cases

Company Research Assistant

Sales prospecting
Business development
Market research
Competitive intelligence
Account planning

Interview Practice Partner

Job interview preparation
Communication skills practice
Performance feedback
Professional development
Skills assessment

License

MIT License - Free to use for learning and portfolio purposes.

Acknowledgments

Built with Google Gemini AI
Powered by FastAPI
Uses Web Speech API

Ready to explore? Check out the individual project READMEs for detailed setup and usage instructions!

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
company-research-assistant		company-research-assistant
interview-practice-partner		interview-practice-partner
.gitignore		.gitignore
HOW_AGENTS_WORK.md		HOW_AGENTS_WORK.md
README.md		README.md
setup.sh		setup.sh
start-all-in-one.sh		start-all-in-one.sh
start-interview-partner.sh		start-interview-partner.sh
start-research-assistant.sh		start-research-assistant.sh
test_setup.py		test_setup.py

Amdev-5/ai-conversational-agents

Folders and files

Latest commit

History

Repository files navigation