AI-Powered Employee Engagement System
Project Highlights
Client Overview
Our client sought to enhance employee engagement by leveraging AI and machine learning to track office behavior, monitor moods, and deliver personalized greetings. The solution also needed to integrate with their HRMS tool to streamline attendance and improve overall employee satisfaction.
Technical Stack
AWS
FFmpeg
Opencv
Pytorch
Tensorflow
Node Js
React
Industry
Human Resources
Project Duration
5 months
Region
India
Challenges
The client needed a system to improve employee engagement, assess moods, and work with their HRMS tool for automated attendance and mood monitoring.
- ➔ Employee Tracking via CCTV
- ➔ Mood Detection
- ➔ Personalized Greetings
- ➔ Tracking Office Absences
- ➔ HRMS Integration
- ➔ Facial Recognition
- ➔ Audio Greetings
- ➔ Integration with HRMS
- ➔ Cloud Infrastructure
- ➔ Real-time Communication
Solution
We developed an AI-powered employee engagement system integrating facial recognition, real-time emotion analysis, greetings, and HRMS integration to streamline attendance and mood tracking.
- ✓We integrated the existing office CCTV system with an AI-based facial recognition model to detect and identify employees as they enter and exit the office.
- ✓The system records the exact time an employee arrives at or leaves the office, feeding this data into the HRMS tool for automated attendance tracking.
- ✓Using facial recognition and emotion analysis, the system tracks the mood of employees when they enter and exit the office.
- ✓The mood data is analyzed to identify patterns that may help HR in understanding employee well-being.
- ✓A smart speaker setup at the office entrance is connected to the AI system. and As the employee is recognized via CCTV, the AI system triggers a personalized greeting (e.g., 'Good morning, [Employee Name]!') using voice synthesis.
- ✓The greeting is tailored based on the employee's mood and arrival time, making it more personalized.
- ✓The system also tracks how many times an employee leaves the office during the day, logging the information along with the detected mood during each exit.
- ✓The HR team can analyze this data to better understand break habits and employee engagement.
- ✓The system is fully integrated with the client's HRMS tool.
- ✓OpenCV and Dlib for facial detection and tracking using CCTV footage. and DeepFace library for real-time emotion analysis, extracting moods such as happy, sad, neutral, etc.
- ✓Custom trained Convolutional Neural Networks (CNNs) for accurate mood detection.
- ✓Utilize the Google Text-to-Speech (TTS) API for voice synthesis to convert text greetings into audio. Integrate a smart speaker with a Raspberry Pi and Node.js to manage audio output seamlessly.
- ✓REST APIs built using Python (Flask) for communication between the AI tool and the HRMS system.
- ✓AWS Lambda for serverless event triggers, ensuring the system is scalable and cost-effective.
- ✓Using AWS S3 for storing CCTV footage and facial recognition data, and AWS Rekognition as a backup for facial recognition in case of local model failure.
- ✓MQTT protocol for low-latency communication between the CCTV system, AI server, and smart speakers.
- ✓WebSockets for real-time employee mood updates and interactions displayed on HR dashboards.
Conclusion
Our AI-driven equine movement and health monitoring system for Molenkoning combines sophisticated video analysis and heart rate monitoring, providing trainers with real-time, non-invasive insights into a horse's gait and overall health. This solution aids in injury prevention, training optimization, and performance enhancement through data-driven feedback. Available exclusively to Molenkoning clients, it utilizes state-of-the-art AI technology to promote horse well-being and streamline training processes. Upcoming features, such as predictive analytics, will further advance equine care.
