MAJOR PROJECT

END TERM REPORT

InterviBot: Ace Every Interview

Cloud based AI Tool

Submitted By

Aryan Ranjan Khushi Nimawat

R2142210948 R2142210955

Under the Guidance of

Dr. Swati Rastogi

_________________
Assistant Professor
AI Cluster, SOCS

Cloud Computing & Virtualization Technology

School of Computer Science

UNIVERSITY OF PETROLEUM AND ENERGY STUDIES

Dehradun-248007
 Index

Contents Page Number

1. Abstract 2

2. Introduction 2

3. Problem Statement 3

4. Literature Review 4

5. Objectives 5

6. Methodology 6

7. System Requirement 8

8. PERT Chart 11

9. Implementation 12

10. Conclusion 15

11. References 17

1
1. Abstract
InterviBot: Ace Every Interview is an innovative cloud-based AI platform designed to enhance
job interview preparation. Leveraging natural language processing (NLP), machine learning
algorithms, and advanced facial and speech analysis, InterviBot offers a dynamic and interactive
interview simulation experience. The platform generates tailored interview questions in real-time
based on user preferences and job roles, evaluates responses using speech and facial expression
analysis, and provides personalized feedback to improve performance. Utilizing scalable cloud
infrastructure, InterviBot ensures efficient data processing and storage, offering users a realistic,
adaptive, and engaging tool to refine their interview skills and boost their confidence for real-
world job interviews.

2. Introduction
In the modern job market, securing a desirable position requires not only strong qualifications
but also the ability to perform exceptionally well during job interviews. The interview process is
often a pivotal factor in hiring decisions, yet many candidates struggle with effective preparation
due to the limitations of traditional methods. These conventional approaches, which typically
involve static practice questions and generalized feedback, may fail to adequately prepare
candidates for the dynamic and often unpredictable nature of real interviews.
InterviBot: Ace Every Interview is designed to transform interview preparation by integrating
advanced artificial intelligence technologies, including NLP, machine learning, facial
recognition, and speech analysis, with an interactive, cloud-based platform. The core innovation
of InterviBot lies in its ability to deliver a highly personalized and realistic interview experience.
By utilizing cutting-edge NLP techniques, InterviBot generates interview questions tailored to
the specific job roles, industries, and skills relevant to each user. This level of customization
ensures that candidates can practice with questions that mirror the actual challenges they may
face in their target positions.
The platform’s real-time question generation is powered by sophisticated NLP techniques, which

2
understand and adapt to the nuances of various job descriptions and industry requirements. Users
receive questions that are contextually appropriate and relevant, enhancing the quality and
effectiveness of their practice sessions.
InterviBot goes beyond traditional text-based feedback by incorporating facial and speech
analysis. This technology evaluates the user's facial expressions and vocal tone during their
responses, providing a deeper insight into non-verbal communication skills, which are critical in
real interviews. Users receive feedback on how their facial expressions, eye contact, and vocal
tones align with effective communication standards in interviews, allowing them to improve both
verbal and non-verbal skills.
InterviBot's cloud-based architecture ensures scalability and accessibility, allowing it to support
many users simultaneously while maintaining high performance. The platform's interactive
interface replicates the real-world interview environment, offering an engaging and immersive
experience that fosters skill development and confidence-building.
By bridging the gap between traditional practice methods and the demands of real interviews,
InterviBot offers a comprehensive and adaptive solution for job seekers. It empowers users to
practice effectively, receive immediate feedback on both verbal and non-verbal cues, and
continuously improve their interview skills, ultimately enhancing their chances of success in a
competitive job market.

3. Problem Statement
Effective job interview preparation is crucial yet challenging for job seekers, as traditional
methods like generic practice questions and mock interviews often lack adaptability and real-
time feedback. These methods fail to provide personalized guidance and do not consider critical
non-verbal communication elements such as facial expressions and vocal tone, which are
essential in interviews. This gap can leave candidates unprepared for real interview dynamics.
InterviBot: Ace Every Interview addresses these shortcomings by providing an AI-driven,
cloud-based platform that dynamically generates tailored interview questions, evaluates
responses with machine learning, and includes facial and speech analysis. By offering
comprehensive feedback on both verbal and non-verbal aspects, InterviBot delivers a more
effective and personalized preparation experience, helping users improve their interview

3
performance and increase their chances of success.

4. Literature Review
1. Emotion and Confidence Classifier Models
Mrs. P. Swetha, B. Gowthami, Ch. Sreemam, and G. Amruth (2023) propose an AI-Based Mock
Interview Evaluator that incorporates emotion and confidence classifier models to enhance the
effectiveness of mock interviews. This system utilizes machine learning algorithms to analyze
video inputs and detect facial expressions and body language, thereby classifying emotions and
assessing the confidence level of interviewees. The evaluative feedback provided by the system
helps candidates improve their non-verbal communication skills, which are crucial in real-world
interviews [1].

2. Multi-Aspect Analysis: Emotion, Confidence, and Knowledge

Another comprehensive approach is presented by Prof. H. R. Agashe and colleagues (2023) in
their work titled AI-Based Mock Interview Evaluator and Analysis. This system not only
evaluates emotions and confidence but also incorporates a knowledge assessment component. By
analyzing the verbal responses and matching them against a database of expected answers, the
system provides a holistic view of a candidate’s performance across different parameters. This
multifaceted analysis enables more targeted feedback, thus enhancing the overall preparation
experience [2].

3. Smart Interview Systems Using AI

The concept of Smart Interviews Using AI is further explored by Aditi S. More, Samiksha S.
Mobarkar, Siddhita S. Salunke, and Reshma R. Chaudhari (2022). Their research emphasizes the
importance of integrating natural language processing (NLP) and machine learning to simulate
realistic interview scenarios. The system they developed can generate dynamic questions tailored
to the candidate's background and field of expertise in real-time, providing immediate feedback
to help them refine their responses and improve their overall interview skills [3].

4. Remote Hiring and AI-Based Interview Systems

4
B. C. Lee (2021) explores the application of AI in remote hiring through the development of an
AI-Based Interview System. This system leverages AI to conduct interviews remotely, which
has become increasingly relevant in the post-pandemic world. The system integrates facial
recognition technology and voice analysis to assess candidates' emotional states and
communication skills during interviews. [4].

5. AI-Driven Mock Interview Platforms

The Skillup Bot, developed by Shashank Rai, Alisha Miranda, Samiya Jagirdar, and Prof. Nidhi
Chitalia (2022), represents another innovative approach in AI-driven mock interview platforms.
This system uses a combination of NLP and AI algorithms to simulate various interview types,
ranging from technical to behavioral. The platform also provides detailed analytics on the
candidate’s performance, allowing users to track their progress over time and focus on areas
needing improvement [5].

6. AI Technology for Smart Interview Systems

Prof. Taware R.K., Bhagyashree Shinde, Nikita Rasal, and Snehal Ghorpade (2021) present a
Smart Interview System Using AI Technology that integrates several AI techniques to provide
a comprehensive interview preparation experience. Their system focuses on the adaptability of
AI algorithms to offer personalized feedback based on the candidate’s performance. This
adaptability is crucial for providing tailored guidance, which is more effective than generic
feedback typically found in traditional mock interview setups [6].

5. Objective
1. Real-Time Response Evaluation with Facial Analysis and User-Friendly Interface:
 User-friendly interface for easy interaction and practice which analyze real-time facial
expression or emotion.
2. NLP for generating contextual interview questions and feedbacks on Speech Analysis:
 Customize questions based on user profiles and job roles, job description and experience.
3. Speech recognition and real time feedbacks on Speech Analysis:
 Provide detailed feedback on tone, pace, clarity and delivery for Improvement.

5
6. Methodology

Figure 1. Software Development Flowchart

1. Local Development:
 Requirement Analysis: Define core features and gather requirements for NLP, facial, and
speech analysis.
 System Design: Plan the local architecture and choose development tools (React, Node.js,
MongoDB, Python, etc.).
 Set Up Local Environment: Install software and set up a local database and server.
 Front-End Development: Create the UI with React, including interview question display

6
and feedback visualization.
 Back-End Development: Develop server-side logic with Node.js and Express; create
APIs for various functionalities.
 NLP/ML Integration: Integrate NLP and ML tools (NLTK, TensorFlow, OpenCV, etc.);
test locally.
 Testing: Perform unit, integration, and user acceptance testing; refine based on feedback.

2. Cloud Integration:
 Architecture Design: AWS services for hosting and processing.
 Cloud Service Integration: Set up cloud databases and storage; integrate AWS services (S3,
EC2, Lambda, Rekognition, Transcribe).
 Data Migration: Move local data to the cloud.
 Testing on Cloud: Test functionality and performance in the cloud; conduct load testing.

7
 Figure 2. AWS Deployment Architecture

3. Deployment:
 Deploy to AWS
 CI/CD Implementation: Set up CI/CD pipelines for automated testing and deployment.
 Monitoring and Optimization: Use AWS CloudWatch for monitoring; optimize resources
for performance and cost-efficiency.

4. Maintenance:
 Regular Updates: Add features, improve functionality, and adapt based on user feedback.
 Performance Monitoring: Regularly review and optimize performance.
 Security and Compliance: Ensure ongoing security and compliance with data protection
regulations.
8
7. System Requirement
Software Requirements
1. Local Development:
 Operating System: Windows, macOS, or Linux
 Programming Languages:
o Front-End: JavaScript (React), HTML, CSS
o Back-End: JavaScript (Node.js), Python (for NLP/ML tasks)
 Frameworks and Libraries:
o Front-End: React, Redux (if needed for state management), Axios (for API calls)
o Back-End: Express.js (for building server APIs), Mongoose (for MongoDB interaction)
o NLP/ML: NLTK, TensorFlow or PyTorch (for machine learning models), OpenCV, and
Dlib (for facial analysis), SpeechRecognition (for speech-to-text conversion)

2. Cloud Integration:
 Operating System: AWS-managed services
 Programming Languages:
o Same as local development (JavaScript, Python)
 Frameworks and Libraries:
o AWS SDK (for integration with AWS services)
o Additional libraries or tools specific to AWS services and cloud functionalities, such as
AWS CLI, Boto3 (Python library for AWS), and Serverless Framework.

3. Databases:
 Local: MongoDB (for local development and testing)
 Cloud: DynamoDB (cloud-based database for scalability and performance)
4. Cloud Services:
 AWS Services:
o Compute: EC2 (Elastic Compute Cloud for scalable computing capacity), Lambda (for
serverless functions)
o Storage: S3 (Simple Storage Service for storing data and media files)
9
 o Database: DynamoDB (NoSQL database service for handling structured data)
o NLP/ML: AWS Comprehend (for natural language processing tasks), SageMaker (for
building, training, and deploying machine learning models)
o Facial and Speech Analysis: AWS Rekognition (for facial analysis), AWS Transcribe
(for speech-to-text conversion)
o Monitoring: CloudWatch (for monitoring application performance and resources)

Hardware Requirements
1. Local Development:
 Computer:
o Minimum of 8GB RAM
o Modern multi-core
o SSD storage (at least 256GB for fast read/write speeds)
 Additional Hardware: Optional based on personal preferences, such as:
o External monitor for dual-screen productivity
o External keyboard and mouse for ergonomic comfort
2. Cloud-Based Deployment:
 Compute Resources:
o EC2 Instances: Select instance types based on application needs and budget (e.g.,
t3.medium for moderate traffic and processing requirements)
o Lambda: Utilize serverless computing resources for event-driven functions and
microservices
 Storage:
o S3 Buckets: For file storage, media handling, and managing large data sets
 Networking:
o AWS VPC (Virtual Private Cloud): For secure networking, isolation of resources, and
access control configurations (e.g., setting up subnets, security groups, and network ACLs).

10
8. PERT Chart

11
9. Implementation

12
13
14
10. Conclusion
All objectives are achieved.

1. Real-Time Response Evaluation with Facial Analysis and User-Friendly Interface. ✔

2. NLP for generating contextual interview questions and feedbacks on Speech Analysis. ✔

3. Speech recognition and real time feedbacks on Speech Analysis. ✔

15
GitHub Link: https://github.com/khushinim/InterviBot-Ace-Every-
Interview.git

16
11. References
[1] Mrs. P. Swetha, B. Gowthami, Ch. Sreemam, G. Amruth, “AI-Based Mock Interview Evaluator: An
Emotion and Confidence Classifier Model,” IEEE Xplore, 2023.

[2] Prof. H. R. Agashe, Dhanashri Anwat, Prajakta Derle, Payal Nagare, Siddhi Dhavale, “AI-Based
Mock Interview Evaluator and Analysis: To Analyze Emotion, Confidence, and Knowledge” IJARSCT
2023.

[3] Aditi S. More, Samiksha S. Mobarkar, Siddhita S. Salunke, Reshma R. Chaudhari, “Smart Interviews
Using AI” VIVA Institute of Technology ISSN 2022.

[4] B. C. Lee, Ph.D. Candidate, Seoul Business School, Seoul School of Integrated Sciences and
Technologies (aSSIST), Seoul, South Korea, “Development of an AI-Based Interview System for Remote
Hiring,” IAEME 2021.

[5] Shashank Rai, Alisha Miranda, Samiya Jagirdar, Prof. Nidhi Chitalia, “Skillup Bot: An AI-Driven
Mock Interview Platform” IRJMETS 2022.

[6] Prof. Taware R.K., Bhagyashree Shinde, Nikita Rasal, Snehal Ghorpade, “Smart Interview System
Using AI Technology” IRJMETS 2021.

17