INDUSTRY INTERNSHIP REPORT

Submitted in
partial fulfillment of requirement for the award of degree of

Bachelor of Technology
in
Data Science

by

Ms. Ritika P. Khadilkar

Industry Guide

Mr. Shashank Surwase

at

Delphi Analytics

Institute Guide

Dr. Chetan Dhule

Associate Professor

May 2025

Department of Data Science, IoT & Cyber Security (DIC)

G H Raisoni College of Engineering
An Empowered Autonomous Institute affiliated to Rashtrasant Tukadoji Maharaj Nagpur University, Nagpur
Accredited by NAAC with “A++” Grade (3rd Cycle)
CRPF Gate No. 3, Hingna Road, Digdoh Hills, Nagpur – 440 016 (INDIA)
T: +91 9604787184, 9689903286, 9921008391 | E: principal.ghrce@raisoni.net | W: ghrce.raisoni.net
 INDUSTRY INTERNSHIP REPORT

Submitted in
partial fulfillment of requirement for the award of degree of

Bachelor of Technology
in
Data Science

by

Ms. Ritika P. Khadilkar

Industry Guide

Mr. Shashank Surwase

at

Delphi Analytics

Institute Guide

Dr. Chetan Dhule

Associate Professor

May 2025

Department of Data Science, IoT & Cyber Security (DIC)

G H Raisoni College of Engineering
An Empowered Autonomous Institute affiliated to Rashtrasant Tukadoji Maharaj Nagpur University, Nagpur Accredited by NAAC with “A++”
Grade (3rd Cycle)
CRPF Gate No. 3, Hingna Road, Digdoh Hills, Nagpur – 440 016 (INDIA)
T: +91 9604787184, 9689903286, 9921008391 | E: principal.ghrce@raisoni.net |W: ghrce.raisoni.net
© GH Raisoni College of Engineering, Nagpur, Year 2025
 Declaration
I, hereby declare that the Industry Internship report submitted herein has been carried out
by me/us in Delphi Analytics towards partial fulfillment of requirement for the award of Degree of
Bachelor of Technology in Data Science. The work is original and has not been submitted earlier as
a whole or in part for the award of any degree / diploma at this or any other Institution
/ University.

I also hereby assign to G H Raisoni College of Engineering, Nagpur all rights under copyright
that may exist in and to the above work and any revised or expanded derivatives works based on
the work as mentioned. Other work copied from references, manuals etc. are disclaimed.

Name of student Mobile No Mail ID Signature

(Other than Raisoni.net)

Ritika P. Khadilkar 9022360906 ritikakhadilkar52@gmail.com

Place :
Date :
 Certificate

The Industry Internship Report entitled as “Transforming EdTech Through Data

Analytics And Automation” carried out under our supervision in Delphi Analytics by Ritika P.
Khadilkar for the award of Degree of Bachelor of Technology in Data Science. The work
submitted is comprehensive, complete and fit for evaluation.

Mr. Shashank Surwase Dr. Chetan Dhule

Industry Guide Institute Guide
Senior Data Analyst Head of Department
Delphi Analytics Department of Data Science,
CSE (IoT) and CSE (Cyber
Security),
G H R C E, Nagpur

Prof. Kamlesh Kalbande Prof. Nekita Chavhan Morris

III Coordinator Head of Department
Assistant Professor Department of Data Science,
Department of Data Science, CSE (IoT) and CSE (Cyber Security),
CSE (IoT) and CSE (Cyber Security), G H R C E, Nagpur
G H R C E, Nagpur

Mr. Gurpal Singh Dr. Sachin

Dean III Cell Untawale
G H R C E, Director
Nagpur G H R C E,
Nagpur
 ACKNOWLEDGEMENT

The successful completion of a project in a set amount of time is a true testament to the
efforts and directions that have been made along the way. Without the help of various
individuals who contributed to this internship project, this project could not have been
achieved. We acknowledge their contributions willingly.

We extend our sincere thanks to our project guide, Associate Professor Dr. Chetan Dhule,
Department of Data Science, IoT & Cyber Security (DIC), GHRCE, Nagpur. His ever-supportive
guidance, valuable feedback, and meticulous supervision were crucial to the successful
execution of our project.

We are also very grateful to Prof. Kamlesh Kalbandhe, Assistant Professor in the Department
of Data Science, IoT & Cyber Security (DIC), for his regular guidance and constructive
suggestions throughout the project process.

We offer our warm appreciation to Prof. Nekita Chavhan Morris, Head, Department of Data
Science, IoT & Cyber Security (DIC), for her constant encouragement and interest throughout
each phase of our study. Her encouragement, advice, and enthusiastic support significantly
motivated us to work very hard.

We also want to thank Dr. Sachin Untawale, G.H. Raisoni College of Engineering, Nagpur, for
encouraging us and granting us the facility and resources for undertaking our work
successfully.
In conclusion, we are thankful to all the teachers and non-teaching staff in the Department of
Data Science, IoT & Cyber Security (DIC), for their indirect and direct guidance, which
enabled us to deliver this project quite effectively.

Regards,
Ritika P. Khadilkar
 Abstract

With the evolving nature of education technology, there is a compelling need for platforms
that can provide bespoke learning experience along with traditional test delivery. Our study
responds to this need by offering an AI-facilitated Learning Management System (LMS) that
seeks to develop an interactive and adaptable learning environment for students, instructors,
and institutional managers. Our approach leverages artificial intelligence to offer customized
learning paths based on individual performance, engagement levels, and behavioral patterns
—features that most traditional LMS platforms such as Moodle often lack. These features
consist of smart feedback mechanisms, content adaptation via automation, and real-time
analytics.

System architecture is designed to be modular, with a frontend built on React to allow a

responsive and user-friendly user interface, and a Node.js-driven backend to offer powerful
RESTful APIs and support for asynchronous operations for optimized data handling. This
technology stack not only enhances system scalability and performance but also allows third-
party educational tools and services to be integrated seamlessly. In addition to this, AI-
powered components like recommendation engines, NLP modules, and predictive analytics
are integrated in order to facilitate dynamic content creation, automate assessment
feedback, and provide actionable insights to instructors and learners. This holistic approach
results in a more engaging, personalized, and data-driven learning experience.

Learning Management System (LMS) is a data-oriented and

modular architecture that integrates backend services powered by Node.js and ClickHouse
database with frontend technologies such as React, Tailwind CSS, and TypeScript.
Data ingestion pipelines, ETL pipelines, dynamic test generation, machine learning-
based feedback question bank creation from textbooks, analysis of student performance,
and visual dashboards built on Grafana are some of the core components of the system. The
solution enables personalized analytics and smart test construction using AI models.
More meaningful engagement of students and more effective academic interventions are
the outcome of the outcomes, which indicate substantial enhancements in test
administration efficiency, dynamic material delivery, and actionable information for teachers.

Our project provides an AI-augmented Learning Management System (LMS) that is set to
offer both routine test administration and personalized learning experiences owing to the
growing need for adaptive and smart education platforms. For dynamically altering material
allocation and suggesting personalized learning pathways, the system combines user
behavior analysis, participation measurements, and educational performance information.
Unlike static Moodle-type LMS platforms, our offering includes AI-driven smart content
suggestions, automatic feedback loops, and real-time analytics consoles. Through machine
learning algorithms driven by past learning data, smart assessment feedback and content
adaptation are applied, enhancing student understanding and teaching effectiveness.

1
The presentation, data, and logic layers of the LMS are isolated by its modular design. The
frontend is built with TypeScript, React, and Tailwind CSS for a responsive as well as
maintainable user interface. For rapid analytics, the backend utilizes ClickHouse and RESTful
APIs with Node.js. Assessment and textbook information are processed using custom ETL
routines. Dynamic question creation and customized content are enabled by AI modules like
recommendation engines and NLP-driven feedback generators. Live insights into system data
and student performance are offered using Grafana dashboards.

The project has effectively created a system that uses automation and data analytics to
revolutionise EdTech and satisfy the evolving needs of educational establishments. This
cutting-edge system is made to adjust to the ever-changing educational environment,
offering a strong foundation that benefits both teachers and students.

In the future, the emphasis will be on using reinforcement learning strategies to further tailor
educational experiences for each student. The system seeks to develop highly customised
learning routes that address each learner's particular requirements and preferences by
utilising these cutting-edge techniques.

Effort will be directed toward developing the system's AI capacity. This will include
developing content generation to create more relevant and interesting learning materials and
expanding academic assistance tools to provide students with comprehensive assistance.
These developments are expected to significantly enhance the learning process by making it
more engaging, effective, and personalized for every student's learning style.

Keywords: Learning Management System (LMS), Customized Learning Path , Real-time

Analytics , Visual Dashboards , Data Handling , NLP Driven Feedback Generators

2
 LIST OF FIGURES

SR. FIGURE NAME PAGE

NO. NO.
1. Figure 1: Company Logo 2

2. Figure 2: Data Flow Pipeline 1

7
3. Figure 3: Personalized LMS Flow 2
0
4. Figure 4: Exam Attendance Analysis 2
3
5. Figure 5: Student Dashboard 2
5
6. Figure 6 : Faculty Dashboard 2
5
7. Figure 7 : Admin Dashboard 2
6
8. Figure 8 : Marketing Dashboard 2
6
9. Figure 9 : My Assignments Page 2
7
10. Figure 10 : Paper Generation and Management Page 2
7

3
 LIST OF TABLES

SR. NO. TABLE NAME PAGE NO.

1 Table 1: Learning Analytics Framework 23

4
 PUBLICATIONS DETAILS

SR. TITLE OF PAPER CONFERENCE CONFEREN STATUS PAGE

NO. NAME CE DATE NO.

1 Transformi -

ng EdTech
Through
Data
Analytics
And
Automatio
n

5
 INDEX

Abstract i
List of Figures iii
List of Tables iv
List of Publications v

CHAPTER 1 : INTRODUCTION TO COMPANY 1

1.1 About the Company 2
1.2 Historical Background 3
1.3 Location 4
1.4 Operational Structure 5
1.5 Vision and Mission of Company 7
1.6 Product Manufactured 8

CHAPTER 2 : CASE STUDY 9

2.1 Introduction 10
2.2 Problem Identification 12
2.3 Objective 14
2.4 Work Carried Out 16
2.5 Solution Provided 21
2.6 Analytics Done 22
2.7 Results and Conclusion 24

CHAPTER 3 : REFERENCES 28

APPENDICES 31

6
 CHAPTER 1
INTRODUCTION TO COMPANY

1
1.1 About the Company

The innovative data science and analytics company Delphi Analytics was established by a
group of forward-thinking graduates from the esteemed Visvesvaraya National Institute of
Technology (VNIT). During their professional experiences, the founders, who have refined
their talents and knowledge at industry heavyweights including Nykaa, Disney+ Hotstar, and
Mars Inc., discovered a crucial market need. They noted that due to resource limitations,
early-stage startups frequently find it difficult to access and utilize advanced analytics, even
in spite of their creative ideas and promise. Delphi Analytics, a business committed to
democratizing data science by offering scalable and affordable solutions catered to the
particular requirements of B2B and e-commerce enterprises, was founded as a result of this
insight.

The fundamental tenet of Delphi Analytics' goal is that advanced analytics techniques and
technologies ought to be available to companies of all sizes, not just big, well-funded ones.
The organization provides an extensive range of services aimed at enabling companies to
leverage data for growth and strategic decision-making. Among these services are data
operations, which include gathering, processing, and managing data; machine learning,
which helps companies create predictive models and learn from their data; artificial
intelligence, which makes it possible to automate complicated processes and create
intelligent systems; and data visualization, which aids companies in effectively
communicating their data insights through dynamic and captivating visual representations.

Delphi Analytics supports its clients' data-driven projects by taking a comprehensive and
multifaceted approach. Offering analytics as a service is the first pillar of this strategy. With
this strategy, clients may obtain sophisticated data analytics capabilities without having to
make large expenditures in infrastructure or in-house knowledge. Businesses can use Delphi
Analytics' analytics services to gain useful insights and recommendations from data analytics
while concentrating on their core business operations.

Delphi Analytics is a dynamic and forward-thinking data science and analytics company
dedicated to enabling companies to fully utilize data. Delphi Analytics is assisting companies
of all sizes and in all sectors in reaching their objectives and reaching their full potential with
its extensive range of services, multifaceted strategy, and unwavering focus on innovation
and excellence. Delphi Analytics is a reliable and esteemed partner to companies on their
data-driven journey, whether it is by giving data science training programs, developing SaaS
tools, providing analytics as a service, or producing data science-based solutions.

Figure 1. Company Logo

2
1.2 Historical Background

Delphi Analytics was founded in 2021 at a critical juncture in the startup scene, when a
number of businesses were going public and the value of data-driven decision-making was
becoming increasingly apparent. The founders, drawing on their vast experience in analytics
leadership positions at well-known companies, noticed a recurring problem: small startups
often encountered challenges because of limited resources and a lack of specialized
knowledge, whereas large corporations were skilled at using data science to drive their
growth.

Motivated by this disparity, the Visvesvaraya National Institute of Technology (VNIT) alums
founded Delphi Analytics in an effort to establish a more fair marketplace. Their goal was to
establish a business that could provide startups with the same superior analytics skills as
those of industry titans, but in a more affordable and easily accessible manner.

Delphi Analytics has been committed from its founding to building a solid foundation that
combines technological expertise with practical application. The company's constant
dedication to innovation, creation of customer-focused solutions, and continuous
improvement of its own tools and processes to satisfy its clients' evolving needs have all
contributed to its success.

Delphi Analytics has become a reliable partner for businesses looking to use data to propel
their growth and success because of its unwavering commitment to excellence. The
organization has helped several businesses overcome their resource constraints and
compete more successfully in their respective sectors by offering customized analytics
services and state-of-the-art solutions.

Delphi Analytics is known for building solid, cooperative connections with its customers,
making sure that every solution is precisely tailored to the particular needs and goals of the
companies it works with. In addition to fostering the company's quick expansion, this client-
focused strategy has strengthened its standing as a major force in the analytics sector,
committed to enabling startups via the intelligent application of data.

3
1.3 Location

SDSM Analytics Pvt. Ltd.

Ground FLOOR, 297, Shankar Nagar Nagpur - 440010
Maharashtra, India

Office Place : 4- Floor of GHRCE, TBIF, Lokmanya Nagar, Hingna, Nagpur -440061

Mode of Internship: Offline

4
1.4 Operational Structure

With 18 committed staff members that bring a plethora of data science, analytics, and
technology knowledge to the table, Delphi Analytics takes pleasure in having a small but
extremely competent workforce. The company's strategic choice to work with a smaller
workforce enables it to maintain agility and responsiveness while guaranteeing that every
customer receives individualized attention and solutions that successfully handle their
particular goals and difficulties.

The operational framework of the business has been carefully planned to promote a culture
of cooperation and creativity. Delphi Analytics makes sure that its team members can
collaborate easily by promoting open communication and the free flow of ideas. This allows
them to use their varied skill sets and viewpoints to create innovative and practical solutions
for clients. In addition to improving the caliber of services rendered, this cooperative strategy
helps the business promptly adjust to the changing demands of its customers and the larger
market.

Cross-functional teams with expertise in critical service areas, such as data operations,
machine learning, artificial intelligence, and data visualization, are at the core of Delphi
Analytics' organizational structure. This organizational structure facilitates the smooth
integration of services, enabling the business to offer clients all-encompassing assistance
throughout their data journey. Delphi Analytics' cross-functional teams collaborate to
provide end-to-end solutions that produce measurable outcomes, starting with the first
phases of data gathering and processing and continuing through the construction of
sophisticated analytics models and interactive visualizations.

In order to guarantee that clients have access to dependable, high-quality data that can
guide their decision-making, Delphi Analytics' data operations team is in charge of gathering,
processing, and managing data. This group collaborates closely with customers to
comprehend their particular data demands and create solutions that are tailored to their
requirements. The data operations team assists clients in realizing the full potential of their
data and gaining insightful knowledge about how their organization is operating by utilizing
the newest technology and data management best practices.

Delphi Analytics is home to a group of highly qualified and seasoned experts in the fields of
machine learning and artificial intelligence who are committed to creating innovative
solutions that let customers take advantage of these game-changing technologies.

The machine learning and artificial intelligence team at Delphi Analytics is at the forefront of
innovation in these quickly developing domains, from developing predictive models that can
predict future trends and behaviours to developing intelligent systems that can automate
complicated activities and procedures. This team creates customized solutions that provide
quantifiable outcomes and promote business expansion by carefully collaborating with
clients to comprehend their particular difficulties and goals.

5
Another important area of Delphi Analytics’ competence is data visualization, where a
committed group of experts create dynamic and captivating visual representations of data.
Through the utilization of cutting-edge data visualization technologies and methodologies,
this team assists customers in successfully and persuasively communicating their data
insights, empowering them to make well-informed decisions and affect significant change
inside their businesses. The data visualization team at Delphi Analytics is dedicated to
helping clients realize the full potential of their data, from creating immersive and interactive
data experiences to creating custom dashboards and reports.

Given the speed at which data science, analytics, and technology are developing, Delphi
Analytics is steadfastly dedicated to lifelong learning and professional growth. Through a
variety of channels, including industry conferences, workshops, online courses, webinars,
and self-directed research, the organization actively encourages its team members to pursue
continuous education and training. Because of this commitment, the team is able to provide
clients with creative and practical solutions by staying up to date on the newest
developments and trends.

Delphi Analytics gives its staff members the freedom to experiment and push the limits of
data analytics by cultivating an environment that values creativity, curiosity, and innovation.
The company's dedication to excellence, teamwork, and ongoing development is the
foundation of its success. Delphi Analytics has established itself as a reliable partner for
companies wishing to use data because to its small but highly qualified staff, emphasis on
open communication, and emphasis on idea sharing. The business stays at the forefront of
data science and technology advancements, offering its clients innovative solutions that yield
quantifiable outcomes.

6
1.5 Vision and Mission of Company

1.5.1 Vision of Delphi Analytics

• Become the industry leader in analytics services globally.

• Offer clients superior analytics solutions with extra value.

• Give customers the ability to:

• Make data-driven, well-informed decisions.

• Use strategic insights to propel corporate expansion.

• Encourage sustainability and success over the long run.

• Establish standards for customer satisfaction and analytics innovation.

1.5.2 Mission of Delphi Analytics

• Provide clients with high-quality technology solutions that are suited to their needs.

• Provide expert outsourcing services that add genuine value to your company.

• technical proficiency in analytics and data science.

• delivery models that are targeted for accuracy and efficiency.

• Connect the dots between cutting-edge data science and the real-world requirements
of startups and mid-sized businesses.

7
1.6 Product Manufactured

Delphi Analytics offers a wide range of goods and services designed to satisfy the various
needs of B2B and e-commerce businesses. Their products are meant to provide all-
encompassing assistance, from setting up data infrastructure to implementing advanced
analytics.

(A) Data Operations :

Delphi Analytics streamlines data infrastructure to promote smooth integration, improve

teamwork, and allow for quick, well-informed decision-making throughout enterprises. They
increase data accessibility and operational efficiency by removing data silos and enhancing
ineffective systems.

(B) Implementation:

The business is excellent at improving user experience, enhancing marketing strategies,

personalizing customer interactions, and turning unstructured customer data into usable
insights. This tactic helps companies reduce client churn and boost conversion rates.

(C) Artificial Intelligence & Machine Learning:

Delphi Analytics uses AI and ML models to generate insightful data, support wise decision-
making, and optimize processes. These advanced, expandable solutions assist customers in
keeping a competitive advantage.

(D) Visualization & Dashboards:

The business creates sophisticated dashboards and dynamic visualizations that help make
data easier to understand, enabling rapid decisions and clear insights. Strategic planning and
operational excellence are supported by these instruments.

(E) Data Science Training Programs:

Delphi Analytics offers training programs that connect the next generation of IT workers with
real-world industry projects because it recognizes the importance of skill development.
Professionals are prepared for data-centric roles and their learning potential is unlocked by
these programs.

(F) Development of SaaS Tools:

In order to provide e-commerce and digital marketing companies with an all-in-one analytics
platform, Delphi Analytics is creating its own SaaS tools. These solutions are made to be
scalable and user-friendly, catering to the particular requirements of smaller companies.

8
CHAPTER 2
CASE STUDY

9
2.1 Introduction

Educational institutions today experience a number of challenges in analyzing performance

data, conducting student tests, and facilitating institutional growth. The manual and non-
uniform nature of the conventional test-formulation and test-evaluation methods often
results in variation in test difficulty, tedious grading processes, and time lapses in providing
students with feedback. These inefficiencies impact learning by hindering the ability of
teachers to offer timely and relevant insights. these issues are exacerbated by the absence of
a centralized infrastructure for the storage and examination of test data. It is hard for
institutions to efficiently gather and analyze data without a single system, which impedes
their ability to make smart decisions and implement changes.

Higher education is being transformed by the integration of Artificial Intelligence (AI) into
Learning Management Systems (LMS), enhancing automation and personalization of learning
processes. AI-driven solutions enable one to create learning paths that are adaptive to the
needs of every individual learner, boosting engagement and outcomes. Real-time feedback
programs and automated marking systems make testing easier, taking some burden off the
teachers and providing students with immediate feedback on their work.

Organizations can make fact-based decisions, predict outcomes, and identify trends with AI-
driven analytics. In addition to making educational processes more effective, this
revolutionary process builds a more dynamic and responsive learning environment. Schools
can overcome traditional barriers and usher in a more successful and creative education by
using AI.

Students have trouble identifying their strengths and weaknesses when they are not
provided with real-time feedback of their performance, thus limiting their ability to improve
learning outcomes. It is challenging for faculty to receive holistic performance data, which
limits their ability to adjust teaching strategies. It is difficult to implement strategic initiatives
and focused enhancements when administrators lack a clear view of faculty and student
performance. In addition, the absence of disciplined marketing insights renders it hard for
institutions to leverage data to build their businesses. A centralized and automated platform
that can speed up test management, provide actionable performance insights, and support
institutional growth through targeted marketing strategies is needed to address such issues.

Higher education is being transformed due to the integration of Artificial Intelligence (AI) into
Learning Management Systems (LMS), enhancing automation and personalization of learning
processes. AI-based solutions allow learning pathways to be created that are adaptive to the
requirements of every individual learner, enhancing participation and outcomes. Automated
grading software and real-time feedback systems make the assessment process easier,
freeing teachers of some of their workload and providing immediate feedback on students'
performance.

10
The Student Test Management Platform addresses the key challenges that schools and
colleges face in conducting tests for students and analyzing performance data. Traditional
test development and assessment methods are often time-consuming, irregular, and
susceptible to human error, leading to inefficient performance monitoring and late feedback.
This technology provides accuracy and uniformity to test management through automation
of generating question papers and data centralization.

The real-time dashboards of the platform provide administrators, teachers, and students
with beneficial information. While the faculty can adjust their instructional practices based
on batch and individual student performance, students can track their performance and
identify areas for improvement. Administrators receive detailed performance reports and a
marketing dashboard that provides analysis of key marketing metrics to help the institute
grow its business.

At all levels of the organization, the platform facilitates informed decision-making, enhances
faculty productivity, and enhances learning outcomes by combining automation, data-driven
insights, and strategic business growth capabilities. The design of a successful, automated
mechanism for managing student exams and assessing performance data in order to
enhance learning outcomes and institutional growth is the primary objective of the Student
Test Management Platform. With an integrated and easy-to-use platform, this project aims
to address inefficiencies of traditional test management methods and provide stakeholders
with immediate data. Application of AI to pre-code the development of a personalized
question bank is one of the prime objectives. The performance of each student will decide
the compilation of the question bank, ensuring that the questions are centered on their
individual areas of learning and improvement. This will maintain the fairness and integrity of
the test and reduce the time and effort required for developing its content.

11
2.2 Problem Identification

2.2.1 Traditional Test Management's Inefficiencies

The entire learning and assessment process is severely hampered by the numerous
inefficiencies that afflict traditional test management systems in educational institutions. The
manual generation of test papers is one of the main problems. Faculty members spend a
significant amount of time creating these examinations by hand, which is a labour-intensive
process that is prone to mistakes and discrepancies. This manual method frequently results
in a lack of standardization and may produce exams that are not representative of the
desired level of difficulty or breadth of the subject matter being evaluated.

The absence of customization in conventional test management systems is another notable

inefficiency. Students' unique learning demands are not met by standardized exam papers. A
one-size-fits-all strategy ignores the fact that every student has distinct strengths and
shortcomings. A mismatch between the test's content and the students' real learning needs
may arise from this lack of personalization, making the evaluation of their knowledge and
abilities inefficient.

Inefficient data management is another critical issue in traditional test management systems.
These systems often lack a centralized database for storing and managing test performance
data. This decentralized approach makes it difficult to access and analyze student
performance data efficiently. Faculty members and administrators may struggle to retrieve
relevant data, leading to delays in identifying areas for improvement and implementing
effective teaching strategies.

The limited availability of real-time insights and analytics further exacerbates the challenges
in traditional test management. Without real-time data, it is challenging for faculty members
and administrators to make data-driven decisions. This lack of timely information can hinder
the ability to identify areas for improvement and implement effective teaching strategies.
Real-time insights are crucial for understanding student progress and making informed
decisions that can enhance the learning experience.

In conventional test management systems, manual review takes a lot of time and is prone to
human mistake, which might result in inconsistent results. By automating the evaluation
process, students receive accurate and consistent feedback while saving time and effort.

2.2.2 Real-time insights and automation are essential.

It is clear that test management and performance analysis require automation and real-time
information. The time and effort needed for test development, evaluation, and data
administration can be greatly decreased by automating important procedures. Faculty and
administrators can use real-time insights to get the information they need to make wise
decisions and put good teaching practices into practice.

12
Faculty members' time and effort spent on test papers can be greatly decreased by
automating the test creation process. By selecting questions based on objective criteria, this
automation can guarantee the tests' fairness and quality. It is possible to create questions
that accurately evaluate students' knowledge and abilities, leading to a more efficient
evaluation of their learning.

The problems of ineffective data management can be solved by centralized data storage
solutions. Test performance data can be efficiently processed and stored by these systems,
guaranteeing prompt and precise information access. Faculty members and administrators
may easily get and analyze data thanks to the structured database, which gives them the
knowledge they need to make wise decisions.

Faculty and administrators can use real-time insights and analytics to get the data they need
to make informed decisions. Faculty members may be able to modify their teaching methods
and give students more focused assistance thanks to these insights. In order to make wise
decisions and put successful institutional growth strategies into action, administrators can
use these insights to monitor student involvement and institutional growth.

Overall educational results and operational efficiency can be improved by combining

automation, real-time analytics, and strategic business research. Through focused marketing
insights, this integration can promote individualized student development, improve
instructional practices, and propel institutional success. All things considered, the platform
can guarantee improved learning results and simplified test administration, offering a more
effective and efficient test management system.

13
2.3 Objective

The creation of an effective, automated system for administering student exams and
evaluating performance data in order to improve learning outcomes and institutional
development is the main goal of the Student Test Management Platform. Through an
integrated and user-friendly platform, this project seeks to solve the inefficiencies of
conventional test management techniques and give stakeholders real-time data.

(A) Personalized Education:

To determine each student's strengths and shortcomings, the AI-powered system will
examine performance data. The algorithm will create a personalized set of questions
cantered on the student's areas of need based on this analysis. With this individualized
approach, every student is guaranteed a special educational experience catered to their own
requirements.

(B) Effectiveness and Saving Time:

The solution drastically cuts down on the time and effort needed to create test content by
automating the question bank generation process. Faculty members can make better use of
their time and resources by concentrating more on teaching and less on administrative
duties.

(C) Fairness and Quality:

By selecting questions based on objective criteria, the AI-driven approach guarantees the
examinations' impartiality and quality. The questions are intended to provide a more
accurate assessment of students' learning by gauging their knowledge and abilities. This
method reduces prejudices and guarantees that every student is assessed fairly.

(D) Constant Enhancement:

Based on continual performance data, the system will update and improve the question
banks over time. By maintaining the questions' relevance and difficulty, this dynamic
technique fosters ongoing enhancements in the learning outcomes of students.

(E) Analysis of Strategic Business:

Strategic business analysis tools will be included into the platform to monitor student
engagement and institutional progress. With the help of these technologies, administrators
will be able to make data-driven decisions and put successful institutional growth strategies
into action.

14
2.3.1 Centralized Data Storage System

For test performance data to be processed and stored effectively, a centralized data storage
system must be developed. For instructors, students, and administrators to have fast and
accurate access to information, this system will make use of a structured database.

(A) Integration and Management of Data:

Test results, student performance indicators, and administrative documents are just a few of
the data sources that will be integrated into the unified data storage system. By ensuring
that all pertinent data is gathered in one location, this integration facilitates management
and analysis.

(B) Effective Data Processing:

The system will process data efficiently, ensuring that it is stored in a structured format that
allows for quick retrieval and analysis. This efficient processing capability enables faculty
members and administrators to access the information they need promptly.

(C) Data Security and Privacy:

Strong security measures will be put in place by the centralized data storage system to
safeguard private student information. To guarantee the security and integrity of the data,
this comprises data protection rules, secure authentication and authorization procedures,
and encryption techniques.

(D) Scalability and Flexibility:

The system is designed to be scalable and flexible, capable of handling large volumes of data
and adapting to changing requirements. This scalability ensures that the system can grow
and evolve with the institution, supporting its long-term data management needs.

15
2.4 Work Carried Out

The Student Test Management Platform was developed using an approach that aims to solve
the shortcomings of conventional test management systems and offer a complete,
automated, and data-driven solution. To accomplish the platform's goals, the methodology
makes use of cutting-edge tools, technologies, and strategies. The main elements of the
methodology, such as the technologies and tools employed and the operation of the system,
are described in the sections that follow.

The integration of AI and machine learning technologies plays a pivotal role in the Student
Test Management Platform. The Mistral API is utilized to generate personalized question
banks by analyzing student performance data through sophisticated machine learning
algorithms. This approach ensures that the questions created are specifically tailored to
target areas where students need improvement, thereby enhancing their learning
experience. Additionally, ChatGPT is integrated into the platform to offer an AI-powered
interaction interface. This feature provides round-the-clock support to students and delivers
personalized study suggestions based on their test performance, thereby fostering a more
engaging and supportive learning environment.

To store and analyse vast amounts of educational data efficiently, the platform uses Click
House, a high-performance columnar database management system. Click House is
especially well-suited for handling test performance data because it is optimized for real-
time analytics and intricate aggregations. In addition, DBeaver functions as a graphical user
interface, providing an easy-to-use platform that facilitates seamless interaction between
developers and analysts and the Click House database, improving accessibility and data
management.

Python is used by the platform for preprocessing, data transformation, and cleaning,
particularly with the Pandas and NumPy modules. In order to maintain data consistency and
integrity, these technologies are crucial for handling duplicate records, outliers, missing data,
and normalization. Apache Airflow is also used for workflow orchestration, which gives the
ETL pipeline automation, fault tolerance, and repeatability. In order to streamline data
processing activities, Airflow DAGs are in charge of scheduling, managing dependencies,
delivering alarms, and logging.

2.4.1 Architecture Design

The Student Test Management Platform's architecture is made to be efficient, scalable, and
modular, guaranteeing reliable data processing, perceptive learning support, and clear
visualizations. By dividing issues into layers, the architecture ensures modularity, security,
and maintainability. The main elements of the architecture and how they work together are
described in the sections that follow.

16
2.4.2 End-to-End Workflow

From raw client-side inputs to user interface interactions and insights creation, the Student
Test Management Platform's system architecture is built to handle the full lifecycle of
educational data. To accomplish the platform's goals, the architecture uses cutting-edge
technologies and tools, is based on a decoupled client-server model, and is backed by
RESTful APIs.

Figure 2 : Data Flow Pipeline

17
Academic timetables, scanned assignment sheets, and raw student test records are gathered
at the start of the procedure. In parallel, an Optical Character Recognition (OCR) module
collects and processes query data. When extracting text from scanned photos or PDFs, this
module is crucial since it makes sure that all pertinent information is recorded and ready for
additional processing.

Information streams enter the Data Engineering and Processing layer after data gathering.
Here, a number of operations, such as cleaning, formatting, and structuring the data, are
carried out using Python-based scripts and Airflow pipelines. The data will be polished and
arranged as a result of this painstaking processing, ready to be entered into the Click House
database. Effective analytical querying is made possible by the database's high-performance
capabilities, which also offer a strong basis for further data processing and visualization.

The Personalized Question Paper Creation module is activated once the structured data is
available. This module creates unique exam papers for every student using large language
models (LLMs) and machine learning models. The module can create questions that are
especially tailored to students' areas of need for improvement by utilizing the Mistral API. By
supporting tailored growth, this customized method makes sure that every student's
particular learning demands are met.

After the data has been analysed, the Dashboard Design and Visualization team uses Grafana
to create aesthetically pleasing dashboards. These dashboards enable administrators,
instructors, and students to make well-informed, data-driven decisions by presenting
performance data in an understandable and intuitive manner. These dashboards' visual
insights are essential for monitoring development, spotting patterns, and putting good
teaching techniques into practice.

The LMS Frontend, where instructors and students engage with the system, consumes all
processed data and insights. The frontend, which was created with React with Tailwind CSS,
provides an engaging and intuitive user experience. This interface is intended for viewing
customized question papers, submitting tasks, and retrieving test results. Third parties, such
mentors or instructors, can also directly enter assignment-related data using the frontend
interface. After processing, this data is saved in the Click House database, which makes
assignment tracking and administration more effective.

2.4.3 Frontend Architecture

The platform's frontend architecture is built with an emphasis on modular, component-

based development, guaranteeing scalability, maintainability, and smooth interaction with
AI-driven features and backend services. Test report visualization, assignment submission,
the display of customized question papers, and efficient user-system interactions are all
18
made possible by this design.

Frontend development uses a technology stack that consists of HTML, CSS, JavaScript, React,
and Tailwind CSS. This cutting-edge and effective stack guarantees that the platform will
continue to be responsive and lightweight while supporting modular user interface
components. A dynamic and eye-catching user interface is made possible by Tailwind CSS's
utility-first framework and Reacts component-based architecture.

The platform guarantees an easy-to-use and captivating user experience by adhering to

modern UI/UX design concepts. Consistency, responsiveness, accessibility, user-cantered
design, feedback, and interactivity are all included in these concepts. In addition to being
feature-rich, this adherence guarantees that the platform is properly aligned with real-world
educational use cases. A smooth and unified user experience is ensured by tailoring the LMS
interface to the client's educational workflow and branding. It offers users a cohesive and
effective experience thanks to its deep integration with the Click House database and
backend services.

2.4.4 Backend Architecture

The platform's backend architecture is intended to act as its structural backbone, supporting
the frontend features, managing essential logic, guaranteeing safe data transfer, and
incorporating AI-based services for astute decision-making. The service-oriented design of
this architecture encourages scalability, modularity, and effective component
communication. Delivering data-driven services and real-time educational insights need it.

Node.js, an event-driven, high-performance JavaScript runtime based on Chrome's V8

engine, is used to create the backend. Node.js was selected because of its non-blocking I/O
style, which enables the platform to effectively manage several concurrent connections.
Applications in real-time schooling require this capacity. A well-organized collection of
RESTful APIs is used by the backend to expose services, guaranteeing stateless
communication, enhancing modularity, and making caching simple. The responsiveness and
scalability of the platform are improved by this configuration.

Business rule processing, user authentication, role-based access management, and the
integration of AI features like performance feedback and automatic question development
fall under the purview of the server-side logic. User identification and authorization, AI
engine integration, assignment management, and dashboard query execution are among the
essential server-side functions. The backend defines routes and middleware for preparing
requests using Express.js, a simple and adaptable Node.js web application framework. The
backend is guaranteed to be reliable, secure, and able to manage intricate business logic and
user interactions with ease thanks to its architecture.

19
The Student Test Management Platform's architecture is made to be efficient, scalable, and
modular, guaranteeing reliable data processing, perceptive learning support, and clear
visualizations. To accomplish the platform's goals, which include increased accuracy and
efficiency, individualized student development, and better teaching techniques, the
architecture makes use of cutting-edge technology and tools. In addition to offering a
thorough and customized approach to student support and institutional growth, the
architecture's tiered design divides concerns and guarantees modularity, security, and
maintainability.

Figure 3 : Personalized LMS Flow

20
2.5 Solution Provided

The difficulties of creating tests by hand were intended to be addressed by the AI-driven
question bank generation system. The system carefully examines student performance data
using machine learning techniques to create customized question banks for every student. By
concentrating on areas that need work and encouraging individualized progress, this method
guarantees that the questions are especially designed to meet each student's unique
learning needs.

This system's capacity to drastically cut down on the time and effort typically needed to
create test content is one of its most notable characteristics. Instead of being bogged down
by administrative duties, faculty members may spend more time teaching and helping
students thanks to this efficiency. Through the application of objective criteria for question
selection, the system's emphasis on quality and fairness is realized, guaranteeing that the
questions fairly evaluate students' knowledge and abilities.

A centralized data storage system was created to address the problems caused by ineffective
data management. To ensure fast and precise information access, this system stores and
manages test performance data using a structured database. The system's effective
processing and storage features make it simple to get and analyse data, making it a useful
tool for administrators, students, and staff.

Information is continuously correct and up to date because to the system's connection with
other parts, like the dashboard visualization system and the AI-driven question bank
generating system. This smooth data transfer between the platform's many components
improves user experience overall and facilitates well-informed decision-making.

The dashboard visualization system was developed to provide administrators, teachers, and
students with insightful information. The system uses Grafana to create aesthetically pleasing
dashboards that display performance statistics in a clear and simple manner. Students may
track their progress and pinpoint areas for development with the help of these dashboards,
which give them individualized performance reports. The reports give a thorough picture of
the students' academic path and provide in-depth insights into their performance,
highlighting both their strengths and flaws.

21
2.6 Analytics Done

Using a range of methods and instruments, the processed data was thoroughly examined in
order to derive insightful conclusions. The main features of the data, including the mean,
median, and standard deviation, were compiled using descriptive statistics. These figures
provided a thorough overview of students' entire performance, emphasizing trends and
patterns that are essential to comprehending their academic trajectory.

To display the performance data in an understandable and intuitive manner, data

visualization techniques were used. The data was represented succinctly and clearly via the
visualizations, which comprised dashboards, graphs, and charts. This method made it easier
to identify patterns and key performance indicators, allowing stakeholders to use the visual
insights to inform their decisions.

Machine learning algorithms played a pivotal role in analyzing the performance data and
identifying areas for improvement. These algorithms were trained on historical data to
predict future performance and provide personalized recommendations. By leveraging the
power of machine learning, the platform was able to offer tailored insights and suggestions,
enhancing the overall learning experience for students.

By identifying areas that needed attention and improvement, the performance data analysis
provided insightful information about student performance. The development of suggestions
for administrators and faculty members benefited greatly from these findings. Targeting
certain areas of growth, personalized learning plans were developed together with
suggestions for further resources like study guides and practice exams. Effective teaching
techniques were also developed for faculty members, including recommendations for
modifying instructional procedures and offering students focused assistance. Additionally,
plans for student involvement and institutional expansion were developed, including
suggestions for focused advertising efforts and student assistance initiatives.

Performance data analysis revealed areas for improvement and offered insightful
information about student performance. These realizations were crucial in creating
individualized lesson plans and successful instructional techniques. They also influenced
plans for student involvement and institutional expansion.

22
 Table 1 : Learning Analytics Framework

Technique/Tool Purpose Outcome

List the main features of the An overview of all students'

Descriptive Statistics data, such as the mean, performance, highlighting
median, and standard trends and patterns.
deviation.

Display performance data in Clear and concise

Data Visualization an understandable manner representation of data,
using dashboards, graphs, identification of key
and charts. performance indicators.

Analyze performance data Personalized advice,

Machine Learning and identify areas for insightful analysis, and ideas
Algorithms improvement, predict future for improving the
performance. educational process.

Develop personalized Targeted areas of

Insights and learning plans, teaching improvement,
Recommendations strategies, and institutional recommendations for
growth strategies. resources, teaching
methods, and support
programs.

23
 Figure 4 : Exam Attendance Analysis

2.7 Results and conclusion

A] Efficiency and Accuracy of Test Management

Student test management was made much more accurate and efficient with the Student Test
Management Platform. Manual jobs used less time and effort when important procedures,
such test creation and evaluation, were automated. While the centralized data storage
system allowed for rapid and precise information access, the AI-driven question bank
generating system guaranteed the exams' fairness and quality.

B] Real-Time Insights and Decision-Making

Faculty and administrators were able to make data-driven decisions thanks to the platform's
24
real-time insights and analytics. Finding trends and patterns was made easier by the
dashboard visualization system's clear and simple presentation of performance data. While
administrators were able to monitor student involvement and institutional progress, faculty
members were able to modify their teaching methods and offer students focused support
thanks to the findings.

C] Personalized Student Growth and Teaching Strategies

By offering customized performance reports and suggestions, the platform facilitated the
progress of each individual learner. Faculty members were able to create efficient teaching
methods and offer focused assistance, while students were able to monitor their
development and pinpoint areas for growth thanks to the insights. Additionally, the platform
gave administrators access to a marketing dashboard and high-level performance reports,
empowering them to make informed decisions and put successful institutional growth
strategies into action.

D] Operational Efficiency and Streamlined Test Management

The automation of key processes and the integration of real-time insights and strategic
business analysis resulted in a more streamlined and efficient test management process. The
platform reduced manual effort and processing time, resulting in a more efficient and
effective test management system. The insights and recommendations provided by the
platform enabled faculty members and administrators to make informed decisions and
implement effective strategies for institutional growth.

The Dashboards Page is the main location for data-driven insights, providing administrators,
instructors, and students with real-time visualizations that help them make decisions based
on engagement and performance indicators.

2.7.1 Student Dashboard

25
 Figure 5 : Student Dashboard

2.7.2 Faculty Dashboard

Figure 6 : Faculty Dashboard

2.7.3 Admin Dashboard

26
 Figure 7 : Admin Dashboard

2.7.4 Marketing Dashboard

Figure 8 : Marketing Dashboard

2.7.5 My Assignments

27
 Figure 9 : My Assignments Page

2.7.6 Paper Generation

Figure 10 : Paper Generation and Management Page

28
 CHAPTER 3
REFERENCES

1. Alhamdi, Asma. (2025). The Role of EdTech in Supporting Lifelong Learning. Journal of
Information Systems Engineering and Management. 10. 439-449. 10.52783/jisei.
v10i12s.1854.

2. Ip, Ken. (2024). The rise of EdTech: Transforming education through entrepreneurial
ventures. Advances in Online Education: A Peer-Reviewed Journal. 3. 177.
10.69554/NSVD4541.

3. Zhao, Xinyu & Ng, Rebecca & Zomer, Chris & Duffy, Gavin & Sefton-Green, Julian. (2025).
RESEARCHING THE EDTECH INDUSTRY FOR CHILDREN: METHODOLOGICAL REFLECTIONS
ON A DESIGN-BASED APPROACH. Air Selected Papers of Internet Research.
10.5210/spir.v2024i0.14080.

4. Kumar, Suresh & Dr, Deshauna. (2024). EdTech in India: Challenges and Opportunities for
Transforming Education in a Digital ERA. 3. 13 - 19.

29
5. Kim, Minyoung & Kim, Ja & Lee, Dae & Woo, Ho & Kim, Yong. (2024). Evaluation Criteria
for Developing a Quality Management System for EdTech Tools. International Journal on
Advanced Science, Engineering and Information Technology. 14. 1199-1208.
10.18517/ijaseit.14.4.16976.

6. R. Paul and M. Rashmi, "Student Satisfaction and Churn Predicting using Machine
Learning Algorithms for EdTech course," 2022 10th International Conference on
Reliability, Infocom Technologies and Optimization (Trends and Future Directions)
(ICRITO), Noida, India, 2022, pp. 1-6, doi: 10.1109/ICRITO56286.2022.9965116.

7. M. Verma, A. Verma and A. Goswami, "Exploring the Adoption of Human Resource

Analytics by Human Resource Professionals of the EdTech Sector," 2024 8th International
Conference on Computing, Communication, Control and Automation (ICCUBEA), Pune,
India, 2024, pp. 1-6, doi: 10.1109/ICCUBEA61740.2024.10774611.

8. A. Birari, P. Yawalkar, R. Janardanan, L. Subramani and R. Sharma, "Perception and

Practices of EdTech Platform: A Sentiment Analysis," 2022 International Conference on
Trends in Quantum Computing and Emerging Business Technologies (TQCEBT), Pune,
India, 2022, pp. 1-8, doi: 10.1109/TQCEBT54229.2022.10041491.

9. N. D. I. Lestari and A. P. Subriadi, "EdTech Investment: Optimism, Pessimism, and

Uncertainty," 2021 International Conference on Electrical and Information Technology
(IEIT), Malang, Indonesia, 2021, pp. 239-245, doi: 10.1109/IEIT53149.2021.9587429.

10. O. A. C. Rojas et al., "Moocs and other technological proposals in the transformation of
education," 2020 15th Iberian Conference on Information Systems and Technologies
(CISTI), Seville, Spain, 2020, pp. 1-7, doi: 10.23919/CISTI49556.2020.9141167.

11. N. Kamal, A. Avasthi, S. Yadav and J. Singh, "Improve Academic Query Resolution through
BERT-based Question Extraction from Images," 2024 IEEE International Conference on
Interdisciplinary Approaches in Technology and Management for Social Innovation
(IATMSI), Gwalior, India, 2024, pp. 1-4, doi: 10.1109/IATMSI60426.2024.10502904.

12. N. Assaf and L. F. Morán-Mirabal, "Instructional Usability and Learner-User eXperience

Assessment in a Virtual Reality Educational Milieu: A Deductive Tech-Instructionality
Model from EdTech," 2023 Future of Educational Innovation-Workshop Series Data in
Action, Monterrey, Mexico, 2023, pp. 1-8, doi: 10.1109/IEEECONF56852.2023.10104873.

13. F. T. Shageeva and I. M. Gorodetskaya, "EdTech competence of engineering university

professors: Research on pshychology and education," 2015 International Conference on
Interactive Collaborative Learning (ICL), Firenze, Italy, 2015, pp. 177-180, doi:
10.1109/ICL.2015.7318021.

30
31
 APPENDICES

A. Group photo with Project Guide Dr. Chetan Dhule

32
B. Group photo with Internship Guide Mr. Shashank Surwase

33
C. Plagiarism Report

34
D. Research Paper

35
36
37
38
39
40