Final Year Internship Report

on

AI/ML APPLICATION DEVELOPMENT

at
Network Software Pvt. Ltd.

Submitted By:
ANSHIKA AGRAWAL
(0901CS221019)

Faculty Mentor:
Prof. Khushboo Agarwal

Assistant Professor

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING

MADHAV INSTITUTE OF TECHNOLOGY & SCIENCE -
DEEMED UNIVERSITY
GWALIOR - 474005 (MP) est. 19

JULY-DEC 2025

0
 PAGE
1 | Page

1
 PAGE
1 | Page

2
 Candidate’s Declaration
I hereby declare that the work presented in this internship report titled “AI/ML
Application Development at Netlink Software Pvt. Ltd.” is an original record of
my work carried out under the guidance of my industry mentor Ms. Soumya Singh
and faculty mentor Dr. Kuldeep Tripathi. The content has not been submitted
elsewhere for the award of any degree or certificate.

Signature: ___________
Anshika Agrawal

PAGE
1 | Page

3
 Acknowledgement
I express my sincere gratitude to Netlink Software Pvt. Ltd. for providing me the
opportunity to pursue my internship. I am deeply thankful to my industry mentor Ms.
Soumya Singh for her continuous guidance and encouragement, and to my faculty
mentor Dr. Kuldeep Tripathi, Department of CSE, MITS Gwalior, for his valuable
academic supervision.
I also thank the Vice Chancellor, Dr. R.K. Pandit, and Head of Department, Dr. Manish
Dixit, for their support. This internship enhanced my technical, analytical, and teamwork
skills, and gave me exposure to real-world industrial projects in Artificial Intelligence
and Data Science.

4
 ABSTRACT
This internship at Netlink Software Pvt. Ltd. provided a valuable
opportunity to gain hands-on experience in the field of Artificial
Intelligence and Machine Learning, focusing on real-world problem-
solving and enterprise-grade solution development. The internship work
revolved around two primary domains: time-series forecasting for financial
data and natural language processing (NLP) for customer interaction
analytics.

The first project involved developing a stock price forecasting and

visualization dashboard. Using the Seasonal Autoregressive Integrated
Moving Average (SARIMA) model, we performed time-series analysis to
forecast short-term stock trends based on historical data. The output was
integrated into an interactive frontend built with [Link], allowing for
rich data visualization, including candlestick charts, moving averages, and
future trend projections. This system serves as a powerful tool for financial
analysts and decision-makers to assess market behavior and adjust
strategies accordingly.

The second project focused on analyzing voice calls from customer support
centers. It incorporated several state-of-the-art NLP and audio processing
techniques such as speech-to-text transcription, speaker diarization using
[Link], and sentiment analysis via VADER. These components
were combined into an end-to-end pipeline capable of generating
meaningful insights from unstructured audio data. The insights included
speaker-level sentiment trends, call summaries, and emotional intelligence
indicators — aiding businesses in improving customer experience and
monitoring support quality.
These projects not only expanded my technical proficiency in AI/ML tools,
Python libraries, and API integration but also enriched my understanding of
how machine learning models are deployed and utilized in business
contexts. Overall, the internship contributed significantly to my
professional development in the fields of data science and applied artificial
intelligence.

5
 Table of Contents

 Candidate’s Declaration
 Acknowledgement
 Abstract
 List of Abbreviations
 Chapter 1: Introduction
 Chapter 2: Company Profile
 Chapter 3: Functioning of Various
Departments
 Chapter 4: Internship Work and Learnings
 Chapter 5: Conclusion
 Bibliography
 Appendices (FPRs, Certificates)

6
 LIST OF ABBREVIATIONS

SARIMA - Seasonal Autoregressive Integrated Moving Average

NLP - Natural Language Processing
VADER - Valence Aware Dictionary and sEntiment Reasoner
ASR - Automatic Speech Recognition
API - Application Programming Interface
Pyannote - Speaker Diarization Toolkit

7
Chapter 1: Introduction
In today’s data-driven world, organizations across industries are leveraging
Artificial Intelligence (AI) and Machine Learning (ML) to extract insights,
forecast trends, and enhance user experiences. The rapid growth of
unstructured data — in the form of text, audio, video, and time-series
signals — has necessitated the development of intelligent systems that can
process, analyze, and act upon such data. The internship at Netlink
Software Pvt. Ltd. provided an immersive experience in designing and
developing such intelligent solutions tailored for enterprise applications.
The internship was structured to expose interns to the complete lifecycle of
data-centric software development — including requirement gathering,
model selection, data preprocessing, algorithm implementation, backend
integration, and visualization. The two main projects carried out during the
internship involved highly practical use-cases: (1) financial forecasting
using time-series models, and (2) sentiment analysis of audio conversations
using advanced NLP and audio processing pipelines.
The first project — focused on stock market forecasting — emphasized the
application of SARIMA models for temporal prediction of market
behavior. We built a web application that not only forecasted stock prices
but also presented them in an interactive format using [Link]. This
introduced a strong link between backend statistical modeling and frontend
data storytelling, both of which are critical for stakeholders in finance.
The second project delved into the domain of speech analytics. With the
proliferation of customer support centers, understanding customer emotions
and agent behavior through voice data has become increasingly vital. By
implementing diarization models to distinguish between speakers and
sentiment classifiers to detect emotions, we were able to provide a
structured view of call dynamics — thereby aiding performance evaluation
and customer satisfaction analysis.
This internship thus served as a bridge between academic knowledge and
industrial application, offering exposure to modern frameworks and
practices used in the AI/ML landscape. It has significantly strengthened my
capabilities in building intelligent systems that derive value from complex
datasets, and has prepared me for a professional career in AI-focused
software development.

8
 Chapter 2: Company Profile
Netlink Software Pvt. Ltd. is a global IT solutions provider headquartered
in Bhopal/Indore. It provides expertise in digital transformation, cloud
services, ERP solutions, and AI/ML development. The company has a
reputation for delivering scalable, data-driven enterprise applications across
industries including BFSI, healthcare, and telecom.

 In 1998, a young technology enthusiast, Anurag Shrivastava, had an

idea and a blueprint to create an IT start-up with a difference. The
vision and passion to build a data-driven platform to generate a breed
of young aspiring IT professionals took shape into Netlink.
 With a mission to help companies reach peak performance with the
technology solutions and services they need and propel innovation
and transformation in business solutions, Netlink has set milestones
one after another and emerged as one of the fastest-growing IT
companies in its areas of operations.
 Headquartered in Madison Heights, US, Netlink has a global
presence. With more than ten offices located at strategic locations,
Netlink serves its clients all over the world.

9
 Chapter 3: Functioning of Various Departments

The company has multiple departments including:

 Software Development & AI Lab – R&D on ML/DL applications.
 Business Intelligence Team – Data visualization and reporting
solutions.
 Cloud & DevOps Team – Infrastructure management and
deployment.
 Client Engagement Team – Requirements gathering and delivery
support.

10
 Chapter 4: Internship Work and Learnings
Project 1: Highcharts Integration for Stock Prediction using SARIMA.

Objective: To forecast stock price trends and visualize results

interactively.

Tools & Technologies: Python (Statsmodels, Pandas), Flask,

[Link].

Methodology: Stock market datasets were preprocessed and modeled using

SARIMA. A REST API was developed with Flask to serve predictions, and
[Link] was integrated to build interactive financial dashboards.

Outcome: Delivered a web-based dashboard with multi-step forecasting

capabilities and dynamic visualizations, enabling improved decision-
making.

Project 2: Call Transcription, Speaker Diarization and Sentiment Analysis

Objective: To extract actionable insights from customer service calls.

Tools & Technologies: [Link], VADER, HuggingFace

Transformers, Python.

Methodology: Audio calls were processed through ASR (Automatic

Speech Recognition) for transcription. [Link] was used to separate
speakers, and VADER sentiment analysis was applied to classify emotions.
The pipeline produced structured reports showing sentiment trends.

Outcome: Provided Netlink with insights into customer satisfaction levels

and agent performance.
Key Learnings
- Exposure to deploying ML models in production.
- Experience in data visualization using Highcharts.
- Practical understanding of NLP and audio analytics.
- Improved teamwork, documentation, and presentation skills.

11
 Chapter 5: Conclusion
This internship offered valuable exposure to AI/ML applications in real-
world scenarios. Through the two projects, I enhanced my expertise in
forecasting, visualization, and speech analytics. The experience
strengthened my ability to integrate research models with industrial needs
and prepared me for future roles in applied AI and data science.

12
 Bibliography
1. Box, Jenkins, Reinsel: Time Series Analysis – Forecasting and Control
2. Highcharts Documentation - [Link]
3. Pyannote Documentation - [Link]
4. VADER Sentiment Analysis - [Link]
5. Scikit-learn Documentation - [Link]
6. Pandas Documentation - [Link]

13