VISVESVARAYA TECHNOLOGICAL UNIVERSITY
"JNANA SANGAMA", MACHHE, BELAGAVI-590018
Industry Internship Report
on
AI/DS at Cambrian SkillsDA & Soil Nutrient Requirement
Prediction
Submitted in partial fulfillment of the requirements for the VIII semester
Bachelor of Engineering
in
Computer Science and Engineering
of
Visvesvaraya Technological University, Belagavi by
Raghav Joshi
(1CD21CS119)
Under the Guidance of
Prof. Arun P
Assistant Professor Dept. of CSE
Department of Computer Science and Engineering
CAMBRIDGE INSTITUTE OF TECHNOLOGY, BANGALORE-560 036
K.R. PURAM, BANGALORE – 560 036, Ph: 080-2561 8798 / 2561 8799
Fax: 080-2561 8789, email: principal@cambridge.edu.in
Affiliated to VTU, Belagavi| Approved by AICTE, New Delhi| NAAC A+ & NBA Accredited|
UGC 2(f) Certified| Recognized by Govt. of Karnataka
2024-2025
CAMBRIDGE INSTITUTE OF TECHNOLOGY
AN AUTONOMOUS INSTITUTION AFFILIATED TO VTU
K.R. Puram, Bangalore-560 036
DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING
CERTIFICATE
Certified that Mr. Raghav Joshi bearing USN 1CD21CS119 , a bonafide student of Cambridge
Institute of Technology, has successfully completed the Industry Internship entitled “AI/DS at
Cambrian SkillsDA Soil Nutrient Requirement Prediction Using Machine Learning ” in
partial fulfillment of the requirements for VIII semester Bachelor of Engineering in Computer
Science and Engineering of Visvesvaraya Technological University, Belagavi during academic
year 2024- 2025. It is certified that all Corrections/Suggestions indicated for Internal Assessment
have been incorporated in the report deposited in the departmental library. The Internship report
has been approved as it satisfies the academic requirements prescribed for the Bachelor of
Engineering degree.
----------------------------- ----------------------------- --------------------------
Internship Guide Internship Coordinator Head of the Dept.
Asst. Prof. Arun P Asst. Prof. Lokesh Dr. Shreekanth Prabhu M Dept. of CSE, CiTech
Dept. of CSE. CiTech Dept. of CSE. CiTech
Name of the Examiners Signature with Date
1.
2.
DECLARATION
I, Raghav Joshi, a student of VIII semester BE, Computer Science and Engineering, Cambridge
Institute of Technology, hereby declare that the Industry Internship has been carried out by me
and submitted in partial fulfillment of the course requirements of VIII semester Bachelor of
Engineering in Computer Science and Engineering as prescribed by Visvesvaraya
Technological University, Belagavi, during the academic year 2024-2025.
I also declare that, to the best of my knowledge and belief, the work reported here does not
form part of any other report on the basis of which a degree or award was conferred on an earlier
occasion on this by any other student.
Date: Raghav Joshi
Place: Bangalore 1CD21CS119
ACKNOWLEDGEMENT
I would like to place on record my deep sense of gratitude to Shri. D. K. Mohan, Chairman,
Cambridge Group of Institutions, Bangalore, India for providing excellent Infrastructure and
Academic Environment at CITech without which this work would not have been possible.
I am extremely thankful to Dr. G. Indumathi, Principal, CITech, Bangalore, for providing me the
academic ambience and everlasting motivation to carry out this work and shaping our careers.
I express my sincere gratitude to Dr. Shreekant Prabhu M, HOD, Dept. of Computer Science and
Engineering, CITech, Bangalore, for his stimulating guidance, continuous encouragement and
motivation throughout the course of present work.
I also wish to extend my thanks to Internship Coordinator, Asst. Prof. Lokesh, Dept. of CSE,
CITech, Bangalore for the critical, insightful comments, guidance and constructive suggestions to
improve the quality of this work.
I also wish to extend my thanks to Internship Guide, Asst. Prof. Arun P, Dept. of CSE, CITech,
Bangalore for the critical, insightful comments, guidance and constructive suggestions to improve
the quality of this work.
Finally to all my friends, classmates who always stood by me in difficult situations also helped me
in some technical aspects and last but not the least, I wish to express deepest sense of gratitude to
my parents who were a constant source of encouragement and stood by me as pillar of strength for
completing this work successfully.
Raghav Joshi
i
ABSTRACT
Soil health is a critical factor in agricultural productivity, directly influencing crop yield and
sustainability. Traditional methods for analyzing soil nutrient requirements are time-consuming,
expensive, and often inefficient. This project focuses on developing a machine learning-based
predictive model to accurately determine soil nutrient requirements based on various soil
parameters such as PH, moisture, organic matter, and macro/micronutrient levels.The proposed
system utilizes supervised learning algorithms, including Decision Trees, Random Forest, Support
Vector Machines (SVM), and Neural Networks, to analyze soil data and predict optimal nutrient
requirements. The dataset comprises historical soil test reports and corresponding fertilization
recommendations. Feature engineering and data preprocessing techniques enhance the model's
accuracy and generalization.By integrating this machine learning model into a user- friendly
application, farmers and agricultural experts can make data-driven decisions about fertilization
strategies, reducing costs and minimizing environmental impact. The project aims to enhance
precision farming, improve soil fertility management, and optimize resource utilization, ultimately
leading to sustainable agricultural practices.
ii
TABLE OF CONTENTS
Chapter Number Content Page Number
COMPANY PROFILE 01
1 1.1 Introduction
01
1.2 Overview of the
Organization 02
2 TASK PERFORMED 04
2.1 Learning Experiences 04
2.2 Knowledge Acquired
04
2.3 Skills Learned
05
2.4 The Most Challenging Task
Performed 05
2.5 Problem Identified
05
3 REFLECTIONS 07
3.1 Solutions 08
3.2 Screenshots
12
CONCLUSION 13
REFERENCES 14
iii
LIST OF FIGURES
FIGURE NO. FIGURE NAME PAGE NO.
3.1 Number Plate Detection System Flow chart 10
3.2 Number Plate Detection system 12
3.2 Number Plate Detection system 12
iv
CHAPTER 1 COMPANY PROFILE
1.1 Introduction
Capgemini is a globally renowned consulting, technology, and digital transformation company
headquartered in Paris, France. Established in 1967, Capgemini expanded its operations to over
50 countries, employing more than 340,000 professionals worldwide. It provides a wide array of
services including consulting, application development, systems integration, cloud services,
artificial intelligence, and business process outsourcing. The company’s vision is to unleash human
potential through technology and innovation. Capgemini partners with businesses across industries
to navigate the rapidly evolving digital landscape and offers tailored solutions that solve complex
problems while boosting efficiency, productivity, and sustainability. Its global delivery model and
strong focus on client-centricity, ethics, and innovation have made it one of the most trusted and
recognized names in the IT and digital services sector.
Capgemini operates with a strong commitment to corporate social responsibility (CSR) and
sustainable development. It has continuously invested in initiatives that foster digital inclusion,
community empowerment, and environmental stewardship. As part of this commitment,
Capgemini collaborates with various non-governmental organizations and social enterprises to
extend the reach of digital transformation beyond corporate walls. One of its most significant
partnerships in India is with the NASSCOM Foundation, a non-profit arm of NASSCOM
(National Association of Software and Service Companies), which is the apex trade body of the
Indian IT and business process management (BPM) industry. Founded in 2001, the NASSCOM
Foundation works towards building a more inclusive and digitally empowered society by
leveraging the power of information and communication technology (ICT) to address pressing
social issues. The foundation’s mission is closely aligned with the United Nations.
The NASSCOM Foundation plays a pivotal role in bridging the digital divide in India by
designing and implementing scalable programs that focus on digital literacy, skill development,
entrepreneurship, and employability. One of the organization’s flagship efforts is the Digital
Literacy Program, which aims to equip marginalized communities—including women, rural
youth, and persons with disabilities—with essential digital skills to navigate today’s world.
Soil Nutrient Requirement Prediction Using Machine Learning Company Profile
B.E., Dept of CSE, CITech 2024-25 Page 1
1.2 Overview of the Organization
Capgemini and the NASSCOM Foundation have developed a strategic partnership that brings
together Capgemini’s technological expertise and NASSCOM Foundation’s deep
understanding of grassroots challenges. Together, they work on creating and executing CSR
programs that focus on increasing digital access and providing opportunities for socio-
economic upliftment through technology. For instance, Capgemini has supported NASSCOM
Foundation’s programs to establish Digital Resource Centres (DRCs) across multiple states in
India. These centres serve as hubs for digital education and career development, especially for
youth and women in low-income communities. Beneficiaries at these centres receive training
in digital literacy, spoken English, financial literacy, and soft skills, often followed by
placement support or entrepreneurship guidance.
Additionally, Capgemini contributes to the foundation’s initiatives aimed at empowering
women entrepreneurs. Under this program, women from rural and semi-urban areas receive
business training, mentorship, and digital tools that enable them to scale their small businesses
and connect to broader markets. These interventions not only foster economic independence for
women but also drive community development, as the income generated is often reinvested in
education, health, and other family needs. Capgemini’s support for such programs exemplifies its
belief in the power of inclusive innovation—using technology not just for profit, but for the
betterment of society.
CHAPTER 2 TASK PERFORMED
2.1 Learning Experiences
Throughout the course of this project, I have undergone a transformative learning experience
that extended far beyond theoretical knowledge. By working on the development of a soil
nutrient requirement prediction system using machine learning, I have come to understand the
powerful role that AI can play in solving critical challenges in agriculture. From the very
beginning, the project immersed me in real-world complexities — handling incomplete, noisy,
and inconsistent soil datasets taught me the practical importance of robust data preprocessing,
feature engineering, and validation. Unlike controlled academic exercises, real agricultural data
demanded iterative cleaning and careful normalization, giving me deep insight into the messy
B.E., Dept of CSE, CITech 2024-25 Page 2
realities of deploying AI in real-world scenarios. Moreover, this project reinforced the
interdisciplinary nature of AI solutions. I had to acquire basic but essential knowledge of soil
chemistry, crop science, and environmental conditions, which directly impact nutrient
availability. I also learned to appreciate user-centric design, particularly as we developed mobile
and web applications for farmers, many of whom face challenges related to digital literacy,
language barriers, and connectivity issues. My experience with cloud computing platforms
demonstrated how modern architectures support scalability and responsiveness, critical for
systems that aim to serve thousands of users across large geographical regions.
2.2 Knowledge Acquired
In terms of knowledge gained, this project significantly deepened my understanding of both
machine learning and soil science. I explored the application of supervised learning algorithms
— including Decision Trees, Random Forest, Support Vector Machines, and Neural Networks
— to predict optimal soil nutrient requirements based on multiple parameters. Alongside these
technical skills, I gained valuable insight into how soil fertility is determined by complex
interactions between pH levels, organic matter, moisture, and macro- and micronutrients like
nitrogen, phosphorus, and potassium. I also acquired an appreciation for the role of satellite-
based remote sensing data and IoT sensor streams in enhancing predictive accuracy. This
knowledge expanded my understanding of how diverse data sources
Soil Nutrient Requirement Prediction Using Machine Learning Task Performed
2.3 Skills Learned
The range of technical and professional skills I developed through this project is extensive. On
the technical side, I gained hands-on experience in data engineering, including cleaning and
preparing complex agricultural datasets for machine learning. I became adept at building and
fine-tuning both regression models, which predict specific nutrient quantities, and classification
models, which categorize soil into nutrient-deficient or nutrient-rich classes. My cloud
computing skills were enhanced as I deployed machine learning models using services such as
AWS SageMaker and Google Vertex AI, ensuring scalability and real-time performance. I also
developed my full-stack capabilities by building web dashboards using React.js and mobile
applications with React Native, both of which provided personalized recommendations to
farmers in a user-friendly format. The project strengthened my skills in integrating IoT devices
B.E., Dept of CSE, CITech 2024-25 Page 3
for real-time data capture and in optimizing system latency to deliver actionable insights within
seconds. Beyond technical abilities, this project improved my competence in technical writing,
project management, and collaborative teamwork, all of which are vital for delivering
successful AI solutions in industry settings.
2.4 The Most Challenging Task Performed
Among the various challenges faced, the most demanding aspect was designing a machine
learning system that could reliably handle real-time, noisy, and dynamic data from multiple
heterogeneous sources. Unlike offline datasets, the IoT sensor data we ingested was prone to
fluctuations due to environmental factors like rainfall, temperature changes, and sensor drift.
Developing data pipelines that could filter out noise while retaining critical patterns was
technically challenging and required careful calibration. Achieving the right balance between
model complexity and real-time responsiveness was another critical hurdle. Deep neural
networks offered superior accuracy but introduced delays that were impractical for field use.
Finding models like Random Forest that could strike the right compromise between speed and
accuracy was a significant achievement. Additionally, designing an interface that was both
simple enough for rural farmers and powerful enough for agricultural experts was a demanding
design challenge, requiring multiple iterations and user testing.
B.E., Dept of CSE, CITech 2024-25 Page 4
Soil Nutrient Requirement Prediction Using Machine Learning Task Performed
2.5 Problems Identified
Through this work, several real-world problems in agricultural soil management became
evident. The foremost issue identified was the inaccessibility of traditional soil testing methods,
which are often expensive, time-consuming, and logistically challenging for smallholder
farmers. As a result, many farmers apply fertilizers based on guesswork, leading to either
overuse, which degrades soil and pollutes water sources, or underuse, which limits crop yield
and economic stability. We also discovered that existing fertilizer recommendation systems are
largely static and fail to adapt to real-time environmental changes such as rainfall and
temperature shifts, which directly affect nutrient availability. Furthermore, we observed a
digital divide among farmers: many lack access to user-friendly digital tools that provide
actionable insights in their local languages and preferred formats, such as voice alerts or SMS
messages. A deeper analysis revealed that current solutions are often fragmented, addressing
either sensor monitoring, machine learning modeling, or mobile application delivery in
isolation. Very few existing platforms offer an integrated, end-to-end precision farming solution
that seamlessly connects soil sensors, predictive models, and intuitive user interfaces. This gap
highlights a significant opportunity for innovation in agricultural technology, which our project
has taken the first steps to address.
B.E., Dept of CSE, CITech 2024-25 Page 5
CHAPTER 3 REFLECTIONS
3.1 Solutions
To address the problems identified during the project, our solution was designed to offer an
integrated, intelligent, and accessible soil nutrient prediction system powered by machine
learning and real-time data sources. At its core, the system leverages supervised learning models
— including Decision Trees, Random Forest, Support Vector Machines, and Neural Networks
— to accurately predict the nutrient requirements of soil based on input parameters such as pH,
nitrogen, phosphorus, potassium, moisture, and organic matter levels. By training on historical
soil test reports and validated fertilization data, the system provides precise, crop-specific
nutrient recommendations, eliminating the guesswork that often leads to over- fertilization or
under-fertilization in traditional farming practices.
One of the key solutions implemented was the real-time integration of IoT sensor data. By
connecting soil sensors directly to our predictive models via cloud infrastructure, we enabled
dynamic recommendations that adapt to changing field conditions such as rainfall, temperature,
and soil moisture variations. This real-time capability ensures that farmers receive timely alerts
and actionable advice, allowing them to adjust their fertilization strategies immediately, rather
than relying on outdated laboratory test results. Additionally, the system was designed to scale
horizontally using cloud platforms like AWS and Google Cloud, making it robust and capable
of serving thousands of farmers across different regions without performance degradation.
To make the solution widely accessible, we developed user-friendly interfaces in the form of a
mobile application and a web dashboard. The mobile app, built using React Native, supports
multiple languages and provides easy-to-understand visualizations, personalized
recommendations, and push notifications. For farmers with limited digital literacy, the app
includes voice-based alerts and SMS notifications, ensuring that critical information reaches
users in formats they are comfortable with. The web dashboard, aimed at agricultural experts
and policymakers, offers deeper insights into soil health trends, enabling data-driven decision-
making at both the micro (farm) and macro (regional) levels.
Soil Nutrient Requirement Prediction Using Machine Learning Reflections
B.E., Dept of CSE, CITech 2024-25 Page 6
3.2 Screenshots
Fig 3.1 Login Page
Soil Nutrient Requirement Prediction Using Machine Learning Reflections
B.E., Dept of CSE, CITech 2024-25 Page 7
Fig 3.2: Prediction
Soil Nutrient Requirement Prediction Using Machine Learning Reflections
B.E., Dept of CSE, CITech 2024-25 Page 8
Fig 3.3: Results
B.E., Dept of CSE, CITech 2024-25 Page 9
CONCLUSION
The Soil Nutrient Requirement Prediction Using Machine Learning project represents a
significant step forward in modernizing agricultural practices through technology. By
leveraging advanced machine learning models, real-time IoT sensor data, and satellite-based
environmental inputs, we have developed a comprehensive, intelligent system that empowers
farmers with accurate, timely, and actionable nutrient recommendations. This solution
addresses the longstanding challenges of traditional soil testing methods — which are often
expensive, slow, and inaccessible — and replaces them with a data-driven approach that is both
efficient and scalable.
Throughout this project, we demonstrated that machine learning models like Random Forest,
SVM, and Neural Networks can effectively predict soil nutrient needs with high precision,
reducing the risks of both over-fertilization and nutrient deficiencies. Our integration of real-
time sensor data and remote sensing technologies ensures that recommendations are dynamic
and responsive to changing field conditions, making the system highly practical for real-world
agricultural environments. Furthermore, by designing a user-friendly mobile application and
web dashboard, we ensured that this advanced technology is accessible to farmers of all
backgrounds, including those with limited technical expertise.
In conclusion, this project has successfully demonstrated the transformative potential of AI in
agriculture. It lays the foundation for future advancements, such as autonomous farming, AI-
driven crop planning, and smart irrigation systems, which will further optimize resource
utilization and environmental conservation. As the system continues to evolve, its scalability
and adaptability ensure that it can become a powerful tool for farmers worldwide, helping them
to maximize agricultural output, safeguard soil health, and contribute to the global challenge
of sustainable food production.
B.E., Dept of CSE, CITech 2024-25 Page 10
B.E., Dept of CSE, CITech 2024-25 Page 11
REFERENCES
[1] "Automatic Number Plate Recognition (ANPR): A Survey" by A. K. Jain, R.
Kasturi, and B. G. Schunck, IEEE Transactions on Intelligent Transportation
Systems, 2021.
[2] "Vehicle Number Plate Recognition using Image Processing and Deep Learning"
by R. Saini and M. Sharma, International Journal of Computer Applications,
2020.
[3] "Real-Time License Plate Recognition Using Deep Learning" by M. Silva and
C. Jung, IEEE Transactions on Intelligent Transportation Systems, 2018.
[4] "YOLO-Based License Plate Detection and Recognition System for Smart
Cities" by K. Verma, A. Srivastava, and S. Gupta, Procedia Computer Science,
2021.
[5] "A Review on Automatic Number Plate Recognition Techniques" by S. A. Jamil
and N. Patel, Journal of Computer Vision and Image Processing, 2020.
[6] "Real-Time Vehicle License Plate Detection System Using CNN and OpenCV"
by M. K. Srivastava, P. Singh, and V. K. Sharma, International Journal of
Advanced Research in Computer Science, 2019.
[7] "Character Recognition in License Plates using Machine Learning Techniques"
by T. Brown and J. Li, Machine Learning and Applications: An International
Journal (MLAIJ), 2020.
[8] OpenALPR – Automatic License Plate Recognition Software.
https://www.openalpr.com
B.E., Dept of CSE, CITech 2024-25 Page
15
B.E., Dept of CSE, CITech 2024-25 Page
