Onboarding Multi - Agent AI Assistant

Employee onboarding platform leveraging Amazon Bedrock Multi-Agent AI, built to boost new hire productivity, minimize manual support workload, and deliver instant and conversational assistance through voice and text.

---

🌟 Overview

An enterprise-grade AI-powered onboarding assistant that reduces time-to-productivity for new joiners, minimizes HR workload through intelligent automation and personalized learning experiences.

Demo Video

https://youtu.be/RX4efKfsTMM

⚙️ Core Components

1. Multi-Agent AI Orchestration - Specialized AI agents working together
2. Conversational Interface - Voice and text-based interaction with emotion detection
3. Document Insights Engine - Extracts knowledge from docs, PDFs, reports
4. Personalized Learning Paths - Adapts to user role, experience, and learning style
5. Automated Workflows - Email automation, scheduling, progress tracking
6. VR/AR Immersive Training - Hands-on learning in virtual environments

---

✨ Key Features

🤖 Multi-Agent AI System

• Orchestrator Agent - Coordinates all specialized agents
• Personalization Agent - Creates adaptive learning experiences
• Content Curator Agent - Manages and recommends content
• Assessment Agent - Intelligent skills evaluation

🗣️ Advanced Voice Intelligence

• Natural language voice interaction with emotion detection
• Amazon Transcribe (Speech-to-Text) + Polly (Text-to-Speech)
• Context-aware responses with conversation memory
• Multi-language support

⚡ Automated Workflows

• Email Automation - Personalized campaigns and notifications
• Smart Scheduling - Automated meeting coordination
• Document Processing - AI-powered knowledge extraction
• Progress Tracking - Continuous monitoring and insights

🎯 Business Impact

• 71% reduction in time-to-productivity
• 85% increase in first-year retention
• 60% decrease in HR support workload
• Real-time insights into onboarding effectiveness

🥽 Immersive VR/AR Training

• Virtual reality office tours and equipment training
• Augmented reality workspace guidance
• Social VR for team building
• WebXR-based (no special hardware required)

Descriptions:

1. Conversational Interface: For voice/text-based interaction
2. Document Insights Engine: Extracts knowledge from KT docs, PDFs, reports
3. FAQ & Troubleshooting Module: Handles repetitive queries and pipeline errors
4. Personalized Assistant Logic: Adapts to user role, team, and tech stack
5. Analytics Dashboard: Tracks usage, gaps, and improvement areas

AWS Services Utilized:

1. Amazon Bedrock Agent: Model access for LLM-driven responses

2. Amazon S3: Stores documentation, reports, and FAQs

3. Amazon Sage Maker ai : For code development and Testing

4. Amazon EC2 : For deploying streamlit application

5. Amazon Polly: Converts text responses to natural-sounding speech

6. Amazon Transcribe: Transcribes user voice input to text

Setup

This Application will run in Amazon sage maker AI and Amazon EC2 as well

1. Configure your AWS credentials:

• o Make sure you have AWS credentials configured with permissions to access Amazon Bedrock

• o You can configure credentials using AWS CLI: aws configure

2. Update the files with Agent Id and Agent Alias id.
3. Running the Voice Agent
4. Run streamlit app using streamlit run voiceagent.py
5. This will automatically call chatagent as well
6. Open your browser and navigate to Local URL
7. You can either speak or chat it will provide options to choose