A

Industrial Training Report

On
Development and Deployment of a Smart Chatbot (AVA) using Machine
Learning and Modern Cloud Technologies
Submitted for the Partial Fulfillment of the Requirement of the
Degree
of
BACHELOR OFTECHNOLOGY
in
CSE(AIML)

Submitted to : Submitted by:

Dr. Chandershekhar Singh Anuj Negi
HOD-Admin AIML/23/039
Dept.of Technology . Enrollment No. 23EJIAI016
JIET Group of Institutions, Jodhpur

Department of Technology
Jodhpur Institute of Engineering & Technology, Jodhpur
2024-25
 CANDIDATE’S DECLARATION

I hereby declare that the work, which is being presented in this Seminar, entitled “Development
and Deployment of a Smart Chatbot (AVA) using Machine Learning and Modern Cloud
Technologies” in partial fulfillment for the award of Degree of “Bachelor of Technology” in
Dept. of CSE with specialization in AIML and submitted to the Department of Technology,
Jodhpur Institute of Engineering and Technology, is are cord of my own work carried under the
guidance of Sourav Soni, Faculty, REGEX Software Services, Jaipur.

I have not submitted the matter presented in this report anywhere for the award of another
degree.

ANUJ NEGI
B.Tech (V Semester)
CSE (AIML)
AIML/23/039

Counter signed by-

Dr. Chandershekhar Singh

HOD-Admin
Dept. of Technology .
JIET Group of Institutions
Jodhpur

i
 CERTIFICATE

REGEX Software Services, (Jaipur)

ii
 ACKNOWLEDGEMENT

The submission of my Industrial Training report, entitled, “CHATBOT AVA USING

MACHINE LEARNING” would not have been possible without the guidance and help of
several individuals who in one way or another contributed and extended their valuable assistance
in the preparation and completion of this seminar. I take this opportunity to express my heartfelt
gratitude towards all of them.

I express my sincerest and utmost gratitude to my Training supervisor Mr. Saurab Soni, Faculty,
REGEX SOFTWARE SERVICES, Jaipur, for his valuable guidance, constant supervision and
continuous encouragement during all the stages of this work . His vast knowledge base,
innovative vision and detailed guidance helped me get through challenges and successfully
complete this training/project.

I wish to express my deep gratitude to Prof. (Dr.) Pratibha Peshwa Swami, Head (Technology)
and Prof. Seminar in - charged Designation, for their valuable guidance and immense assistance
in preparation of this report.

Lastly I would like to express my sincere gratefulness to GOD, my parents and my dear friends
and all those people who have helped me directly or indirectly for the completion of this work.

Date: ANUJ NEGI

Place: Jodhpur B.Tech. (CSE(AIML))

Roll No. : 23EJIAI016

iii
 ABSTRACT

This project presents the design and development of Chatbot AVA, an AI-powered virtual
assistant capable of performing intelligent conversations and assisting users with daily tasks. The
chatbot is developed using Machine Learning and Natural Language Processing (NLP)
techniques to understand user inputs, generate context-aware responses, and deliver a natural
conversational experience.

The front-end interface is built using Streamlit, providing a clean, interactive UI. The backend
integrates Google Generative AI for enhanced NLP capabilities and speech recognition/text-to-
speech for voice interaction. Deployment is done using Docker containers, ensuring portability
and scalability, and hosted on AWS EC2 for 24/7 accessibility.

AVA can assist users by:

 Answering general queries

 Playing mood-based music via Spotify API

 Showing location-based information with maps

 Handling voice-based input/output

 Engaging in personalized interactions

The project bridges the gap between academic AI concepts and real-world application by
combining multiple modern technologies into a single intelligent assistant.

iv
 List of Figures

Fig. No. TITLE Page No.

Fig. 4.1 Modules Implemented 12

Fig. 4.2 Deployment Module 13

Fig. 4.3 Architecture Overview 14

Fig. 5.1 API Integration 18

Fig. 5.2 Comparative Analysis with Industry Projects 23

v
 List of Tables

Table No. TITLE Page No.

Table 5.1 Comparative Analysis with Industry Projects 22

vi
CONTENTS Page No.
Candidate’s Declaration i
Certificate ii
Acknowledgement iii
Abstract iv
List of Figures v
List of Tables vi
CHAPTER 1: INTRODUCTION 1-3
1.1 Background of The Company 1
1.2 Organizational Structure 1
1.3 Nature of Work/Training Assigned 2
1.4 Objectives of the Training 2
1.5 Scope of the Project 2-3
CHAPTER 2 : COMPANY INFRASTRUCTURE 4-6
2.1 Departmental Structure 4
2.2 Network Structure 4
2.3 Infrastructure Facilities 5
2.4 Software Stack Availability 6
2.5 Learning and Work Culture 6
CHAPTER 3: TRAINING ATTENDED 7 – 10
3.1Introduction 7
3.2 Exposure Level 7-8
3.2.1 Technical Exposure 8
3.2.2 Deployment Exposure 8
3.2.3 Project Development Workflow 8
3.3 Tasks Assigned 9
3.3.1 Phase 1: Setup and Learning 9
3.3.2 Phase 2: Chatbot Development 9
3.3.3 Phase 3: Feature Integration 9
3.3.4 Phase 4: Testing and Deployment 9
 3.4 Learning Curve 10
3.5 Team Collaboration and Mentorship 10
3.6 Summary of Outcomes 10
CHAPTER 4 : SYSTEMS/PROJECT DEVELOPMENT 11 – 16
4.1 Project Description 11
4.1.1 Modules Implemented 12
4.1.2 Architecture Overview 12-13
4.2 Role & Responsibilities 14
4.2.1 Development Responsibilities 15
4.2.2 Integration Responsibilities 15
4.2.3 DevOps Responsibilities 15
4.2.4 Team & Communication 16
CHAPTER 5 : CONCLUSION 17 – 25
5.1 Lessons Learned Skills Developed 17
5.2 Knowledge Gained 20
5.3 Career Impact and Future Scope 21
5.4 Comparative Analysis with Industry Projects 22
5.5 Technical Challenges Faced & Solutions 23
5.6 Code Documentation and Versioning 24
5.7 Project Evaluation & Feedback 25
5.8 Final Reflection 25
REFERENCES 26
CHAPTER 1
INTRODUCTION

1.1 Background of the Company

Regex Software Services is a Jaipur-based technology company focused on providing advanced
industrial training and custom software development services. The company specializes in
modern technologies such as Machine Learning, Artificial Intelligence, Cloud Computing,
Full Stack Web Development, and DevOps.

With a strong emphasis on practical, hands-on learning, Regex has partnered with various
educational institutions to help bridge the gap between academic learning and real-world
industry requirements. Their training modules are industry-oriented, often resulting in students
developing complete software projects by the end of their internships.

1.2 Organizational Structure

Regex Software Services operates with a lean yet effective organizational structure. It includes:

 Training Department – Handles industrial and technical training programs for students
and professionals.
 Development Team – Works on client-based projects and research & development
using modern Software stacks.
 Admin & HR Team – Manages student onboarding, project assignments, and
evaluation.
 Mentors/Trainers – Professionals with expertise in Python, AI/ML, DevOps, and
Full-Stack technologies.
 Project Coordinators – Bridge communication between mentors and trainees, ensuring
smooth workflow and timely project completion.

The company follows a collaborative work environment where interns interact directly with
industry experts and receive continuous mentoring.

1
1.3 Nature of Work/Training Assigned
During the training, students are given hands-on experience in live projects or simulated
industry-grade applications. For this project, the nature of work involved:

 Learning and implementing AI/ML models using Python

 Developing a Streamlit-based chatbot interface

 Integrating Google Generative AI, Speech APIs, and Spotify API

 Handling user authentication, file I/O, and voice input/output

 Deploying the chatbot using Docker and hosting it on AWS EC2

This practical exposure helped trainees understand real-world project development cycles,
including design, coding, debugging, and deployment.

1.4 Objectives of the Training

 To apply academic knowledge in a real-world software development environment.

 To develop an intelligent AI-powered chatbot using modern tools.

 To understand the working of voice-based NLP systems and recommendation engines.

 To gain hands-on experience with DevOps tools like Docker and cloud platforms like
AWS.

 To enhance skills in Python, Machine Learning, APIs, and deployment strategies.

1.5 Scope of the Project

The chatbot developed during this training can serve multiple purposes, such as:

 A personal assistant capable of interacting with users through voice/text.

 A music recommendation engine based on mood using Spotify API.

 A location-aware assistant providing real-time information using mapping tools.

2
 A modular platform that can be extended to handle tasks like reminders, weather updates,
news summaries, and more.

 A deployable product on cloud environments that demonstrates AI integration, voice

interaction, and real-time user support.

3
CHAPTER -2
COMPANY INFRASTRUCTURE

2.1 Departmental Structure

Regex Software Services maintains a streamlined yet dynamic departmental structure to support
both student training and industrial project development. The departments operate with strong
interconnectivity, fostering collaboration between trainees, mentors, and real-world clients.

 Training Division: This division is responsible for conducting internship sessions,

delivering curriculum-oriented courses, and assigning real-time projects. It includes
senior mentors, guest instructors, and evaluators.

 Development & R&D Department: Works on cutting-edge technologies like AI/ML,

full-stack development, and automation. Trainees often get to collaborate on these
projects to gain live exposure.

 DevOps & Deployment Department: Handles cloud integration, containerization,

version control, and CI/CD pipelines using tools like Docker, AWS, and GitHub.

 Project Management & Support: Ensures smooth flow of tasks, team coordination,
performance tracking, and reporting. It includes project managers, guides, and reviewers.

This structure ensures that trainees are not just trained, but also experience the lifecycle of a real
software project in an industry-like setting.

2.2 Network Structure

The network infrastructure at Regex Software Services is designed to ensure secure, high-speed,
and collaborative working conditions for students and developers.

4
Key highlights:
 LAN & High-Speed Wi-Fi: The facility is equipped with fiber-optic internet
connections with backup support, ensuring uninterrupted access to cloud servers,
repositories, and APIs.

 Cloud Access: Trainees are given access to tools like AWS EC2, GitHub, and
DockerHub through secure credentials.

 Local and Remote Servers: Projects are developed locally and then tested on remote
EC2 instances via SSH for deployment.

 Firewall and Access Control: To ensure cybersecurity, firewalls are enabled and access
rights are provided based on trainee roles (read-only, contributor, admin).

 Code Collaboration Tools: Platforms like GitHub, Trello, and Slack are used to
promote team communication and version control.

2.3 Infrastructure Facilities

Regex Software Services provides a professional and modern infrastructure conducive to both
learning and development.
Some of the key facilities include:

 Fully Equipped Labs: Systems with the latest software installations (Python, Docker,
Visual Studio Code, etc.) and microphone/speaker support for AI voice projects.
 High-Speed Internet: Dedicated leased line with failover ensures uninterrupted
connectivity during API-intensive training and deployments.
 AI Training Rooms: Designed for ML/AI experimentation and integrated with cloud
platforms.
 Deployment Room: Equipped with local servers and Docker environments for testing
containerized apps.
 Seminar and Review Halls: Used for project presentations, code walkthroughs, and live
debugging sessions with trainers.

5
These infrastructure capabilities ensure students experience a simulation of working in a top-tier
tech environment.

2.4 Software Stack Availability

To support diverse projects, Regex Software Services provides trainees with access to a wide
range of software and platforms:
 Programming Languages: Python, JavaScript, HTML/CSS

 Web Frameworks: Streamlit, Flask, Django

 APIs & AI Tools: Google Generative AI, Spotify API, OpenAI, HuggingFace

 DevOps Tools: Docker, AWS EC2, Git/GitHub, DockerHub

 Database Tools: SQLite, MySQL, Firebase

 IDE/Editors: Visual Studio Code, Jupyter Notebook

 Voice/Audio Libraries: pyttsx3, speechrecognition, gTTS

This software stack enables the implementation of full-stack AI projects, including frontend,
backend, APIs, and cloud-based deployment.

2.5 Learning and Work Culture

Apart from technical infrastructure, the work environment at Regex promotes active learning and
collaboration:

 Daily Standups and code reviews mimic the agile software development model.

 Mentor Sessions are conducted to address doubts and explain core logic.

 Hands-on Assignments ensure every student applies what they learn.

 Peer Learning is encouraged through team-based projects and shared GitHub repos.

 Continuous Feedback helps trainees improve code quality, documentation, and

presentation skills.

6
CHAPTER –3
TRAINING ATTENDED

3.1 Introduction
As a part of the B.Tech curriculum, industrial training plays a vital role in providing students
with hands-on experience and an opportunity to work with current industry tools, environments,
and expectations. I undertook my industrial training at Regex Software Services, Jaipur, for a
duration of 45–60 days, where I worked on a live project titled “Chatbot AVA using Machine
Learning.”

The training was designed to bridge the gap between academic knowledge and real-world
technical applications. It involved a comprehensive curriculum that combined technical theory,
real-time project work, and guided mentorship. The experience helped me strengthen my
understanding of software development life cycles, machine learning workflows, deployment
using cloud platforms, and the integration of third-party APIs for building smart systems.

This period also enabled me to gain insights into professional coding practices, teamwork,
documentation standards, time management, and agile development processes—all under a
professional mentor-supervised environment.

3.2 Exposure Level

During the training, I was exposed to several technologies, tools, and methodologies, many of
which are currently trending in the industry. This practical exposure contributed significantly to
my technical skillset and confidence in developing real-world software systems.

3.2.1 Technical Exposure

 Programming: Python was used as the core language for backend development, data
processing, and API integration.

7
  Frontend Interface: Developed using Streamlit, enabling fast UI rendering and easy
deployment of machine learning models in web apps.

 Artificial Intelligence: Learned how Natural Language Processing (NLP) works

through integration with Google Generative AI API.

 Voice Integration: Implemented speech recognition and text-to-speech for hands-free

interaction using speechrecognition, pyttsx3.

 Spotify API: Used to recommend songs dynamically based on detected user mood.

 Location Integration: Implemented interactive maps using folium and geopy for real-
time user location display.

3.2.2 Deployment Exposure

 Docker: Learned to containerize the entire project using Docker, which makes the project
scalable and portable across environments.

 AWS EC2: Gained exposure to launching cloud servers, deploying the application
remotely, and managing public access to the chatbot system.

3.2.3 Project Development Workflow

 Worked in agile-based sprints, where tasks were assigned weekly.

 Participated in code reviews and debugging sessions with senior mentors.

 Gained familiarity with version control through Git and GitHub.

 Learned about project documentation, commenting standards, and collaborative code

practices.

8
3.3 Tasks Assigned
The training involved structured tasks designed to gradually build up the project. Each phase
included planning, coding, testing, and deployment.

3.3.1 Phase 1: Setup and Learning

 Environment setup (Python, Streamlit, Docker)

 Introduction to APIs and chatbot architecture

 Studied voice I/O libraries and NLP basics

3.3.2 Phase 2: Chatbot Development

 Designed the chatbot interface using Streamlit

 Integrated Google Generative AI API for dynamic responses

 Added voice input/output capabilities

3.3.3 Phase 3: Feature Integration

 Integrated Spotify API to play mood-based songs

 Added location and map functionality using folium

 Implemented error handling and user-friendly messaging

3.3.4 Phase 4: Testing and Deployment

 Dockerized the entire application

 Deployed and tested on AWS EC2 instance

 Collected feedback, fixed bugs, and improved UX

9
3.4 Learning Curve
The training helped me transition from theoretical understanding to practical application. Some
key learnings include:

 AI/ML Implementation: Gained practical knowledge of integrating AI models into

applications, instead of just learning algorithms on paper.
 API Integration: Learned how real-time data can be fetched and manipulated using
APIs.
 Cloud Deployment: Understood how to host a complete project in a cloud environment
securely and efficiently.
 Debugging Skills: Got hands-on experience in finding and fixing bugs in large, multi-file
applications.
 Professional Work Culture: Learned communication, time management, reporting, and
documentation—important soft skills in a professional setting.

3.5 Team Collaboration and Mentorship

 Regular check-ins with mentors for code review and project guidance.

 Collaborative work using GitHub repositories shared among interns.

 Peer-learning environment where fellow trainees helped improve solutions.

 Final project presentation and evaluation by company panel.

3.6 Summary of Outcomes

 Successfully developed a voice-enabled AI chatbot that responds intelligently and
performs real-time tasks.

 Understood end-to-end project architecture: development → testing → deployment.

 Improved technical communication, documentation, and debugging skills.

 Gained confidence in handling real-world projects independently.

10
CHAPTER –4
SYSTEMS / PROJECT DEVELOPMENT

This chapter explains the entire development process of the project titled “Chatbot AVA using
Machine Learning”. It covers the system's purpose, architecture, components, technologies
used, and the roles undertaken during its development.

4.1 Project Description

The AVA Chatbot (AI Virtual Assistant) is a smart conversational agent developed using
Python, Machine Learning, and Natural Language Processing. It is designed to assist users
through interactive text and voice commands, providing intelligent responses and performing
utility tasks like playing music, showing maps, or giving recommendations.

The system is a combination of:

 Frontend UI (Streamlit): Handles user interaction through voice or text.

 AI/NLP Integration: Uses Google Generative AI API to understand and respond to

user queries.

 APIs for external functionalities like Spotify (for music) and Folium (for maps).

 Voice Support: Allows both speaking to and hearing from the bot using
speechrecognition and pyttsx3.

 Cloud Deployment: Fully containerized using Docker and deployed to AWS EC2 for
online access.

4.1.1 Modules Implemented

The system consists of the following core modules:

11
 Deployment Module Google Generative

UI Interface

Music Recommendation

Map Module

Voice Interaction

Fig. 4.1 : Modules Implemented

a) User Interface (UI)

 Built using Streamlit, which renders a simple and interactive web interface.

 Accepts voice and text inputs from the user.

 Displays bot responses, audio playback, and map outputs.

b) Voice Interaction Module

 Uses the speech recognition library to capture user speech and convert it into text.

 Uses pyttsx3 or gTTS to convert bot responses into speech.

 Adds accessibility and realism to the chatbot interaction.

c) Google Generative AI Module

 Sends user queries to Google's NLP API.

 Returns intelligent, context-aware responses.

 Supports both open-ended and task-specific queries.

d) Music Recommendation Module

 Integrates with Spotify API.

12
  Recommends and plays songs based on user mood or request.

 Dynamically fetches song titles, artists, and playback links.

e) Map and Location Module

 Uses folium and geopy to detect and display the user’s location.

 Provides map-based outputs in the Streamlit app.

 Can be extended for location-based services like weather or nearby places.

f) Deployment Module

 Entire project is packaged into a Docker container.

 Deployed on AWS EC2 with a public IP for live testing.

 Maintains environment portability and reduces manual configuration.

Fig. 4.2 : Deployment Module

13
4.1.2 Architecture Overview
The system follows a modular, layered architecture:

Fig. 4.3 : Architecture Overview

1. Input Layer:
o Accepts user input (text or voice).

2. Processing Layer:
o Processes the input using NLP and maps it to a corresponding function.

3. API Integration Layer:

o Communicates with external APIs (Google, Spotify, Maps).

4. Response Generator Layer:

o Creates a response text and speech output.

5. Output Layer:
o Displays the final output on the Streamlit UI.

o Plays audio or renders maps as needed.

14
Each module is loosely coupled and can be extended or replaced without affecting the entire
system.

4.2 Role & Responsibilities

As the primary developer of the Chatbot AVA system during my training, I undertook several
responsibilities that spanned across different domains of software development.

4.2.1 Development Responsibilities

 Designed and implemented the entire Streamlit interface for chatbot interaction.

 Integrated Google Generative AI API to handle intelligent response generation.

 Implemented speech input/output features using Python libraries.

 Wrote modular, reusable code using Python best practices and documentation.

4.2.2 Integration Responsibilities

 Integrated Spotify API for mood-based music recommendation.

 Configured map-based location features using Folium and Geopy.

 Linked various modules through proper API routing and error handling.

4.2.3 DevOps Responsibilities

 Containerized the entire project using Docker.

 Deployed the application on AWS EC2 and managed public access settings.

 Handled environment variables securely using dotenv.

 Managed version control using GitHub.

15
4.2.4 Team & Communication
 Participated in regular mentor meetings and peer reviews.

 Documented the full development process and created presentation materials.

 Contributed to discussions on possible feature extensions and improvements.

16
CHAPTER – 5
CONCLUSION

5.1 Lessons Learned & Skills Developed

The 45–60 days of industrial training at Regex Software Services proved to be a transformative
experience in my academic journey. It provided me with the exposure, hands-on practice, and the
confidence required to apply theoretical knowledge in real-world scenarios.

🔹 Technical Skills Developed

1. Python Programming (Advanced)

o Writing modular, clean, and scalable code using best practices.

o Working with external libraries such as requests, streamlit, folium, speech

recognition, pyttsx3, and dotenv.

2. Streamlit Application Development

o Designing interactive user interfaces with widgets.

o Managing real-time user interactions and handling asynchronous calls.

3. API Integration

o Consuming RESTful APIs like Google Generative AI, Spotify, and Open
Weather.

o Understanding authentication using API keys, OAuth2, and tokens.

17
 Fig. 5.1 : API Integration

4. Voice Processing

o Capturing and converting voice to text (Speech Recognition).

o Generating natural voice responses (pyttsx3, gTTS).

5. Map & Location Processing

o Using Geopy for geolocation and Folium to render maps.

o Displaying current location and mapping it with the application.

6. Containerization with Docker

o Writing Docker files to containerize Python applications.

o Managing image builds and running the app on any OS without setup issues.

18
 7. Cloud Deployment on AWS

o Launching and configuring EC2 instances.

o Hosting the chatbot on a live server using secure credentials and port
management.

8. Version Control with Git & GitHub

o Pushing code commits, collaborating on a shared repo, and using GitHub Issues
for task tracking.

🔹 Soft Skills Developed

1. Project Planning & Time Management

o Breaking down development tasks into weekly milestones.

o Meeting deadlines and submitting working builds for review.

2. Team Collaboration & Communication

o Participating in mentor check-ins and discussing issues openly.

o Documenting code and features to help others understand the project flow.

3. Debugging & Problem Solving

o Diagnosing errors using logs, test prints, and exception handling.

o Exploring alternative solutions and fallback methods.

19
 4. Professionalism

o Following guidelines and protocols set by the company.

o Maintaining proper communication via Slack, Google Meet, and email.

5.2 Knowledge Gained

The practical training environment gave me deep insights into how industry-standard projects are
built, tested, deployed, and maintained.

🔹 Real-World Project Development Lifecycle

 From initial planning, writing features, testing modules, to final deployment, I learned the
importance of maintaining a structured development process.

 Understanding that project development is not only about writing code but also about
writing maintainable, documented, and testable code.

🔹 AI + API Collaboration

 This project taught me how modern chatbots are much more than simple rule-based bots.

 I understood how to leverage large language models like Google Generative AI to

deliver intelligent and relevant responses, personalized to the user’s context.

🔹 Modular Architecture

 Each function in the AVA chatbot — speech recognition, NLP, map display, or music
recommendation — was developed independently, making the system scalable and easy
to debug.

20
  The knowledge of structuring applications into modules is something I can carry into
larger future projects.

🔹 Hands-on with Cloud Technologies

 I now know how to use Docker and AWS EC2 — two of the most demanded DevOps
tools in today’s software industry — to deploy my own applications and share them with
the world.

🔹 Working Under Realistic Deadlines

 Unlike classroom projects, training timelines at Regex were strict and deadline-
oriented.

 This taught me how to prioritize, optimize my code, and deliver features that are stable
and functional within time constraints.

5.3 Career Impact and Future Scope

This training gave me a clearer vision of my career path. I now have a strong foundation to
pursue fields such as:

 Machine Learning Engineer

 AI Chatbot Developer

 DevOps Engineer

 Full Stack Python Developer

 Cloud Deployment Specialist

21
🔹 Future Enhancements to the AVA Chatbot

 Add support for multilingual conversations.

 Integrate user profile memory to personalize recommendations.

 Incorporate calendar/reminder features via Google Calendar API.

 Extend to mobile platforms using React Native + API bridge.

 Add more emotion-based features like sentiment detection.

This chatbot is not a finished product — it is a scalable framework that can grow into a full AI
assistant platform.

5.4 Comparative Analysis with Industry Projects

During the training, we reviewed how actual AI-driven chatbots (like Google Assistant, Siri, or
Alexa) are built in the industry. While our project was academic in nature, many similarities
emerged:

Industry Chatbots (Siri, Google

Feature/Aspect AVA Chatbot Project
Assistant)
NLP Engine Google Generative AI Proprietary ML models (BERT, GPT)
Via SpeechRecognition &
Voice Support Inbuilt voice processing APIs
pyttsx3
Music
Spotify API integration Apple Music, YouTube Music
Recommendation
Scalable microservices on Google
Deployment Docker + AWS EC2
Cloud / AWS
Input/Output Mode Text + Voice Voice-first, multi-modal

Extensibility High – Modular Code Limited (Closed source)

Table 5.1 : Comparative Analysis with Industry Projects

22
This comparison helped me understand what additional features are required to make AVA a
market-ready product and the kind of infrastructure that supports large-scale voice assistants.

Fig. 5.2 : Comparative Analysis with Industry Projects

5.5 Technical Challenges Faced & Solutions

Here’s a breakdown of real-world technical problems I faced and how I overcame them:

🔸 Challenge 1: API Response Latency

 Issue: Google Generative AI API had delayed responses during high load.

23
  Solution: Implemented asynchronous calls in Streamlit using background threading and
spinners to improve UX.

🔸 Challenge 2: Voice Not Working on Some Devices

 Issue: pyttsx3 behaved inconsistently on Linux vs Windows.

 Solution: Added a fallback using gTTS (Google Text-to-Speech) with MP3 playback.

🔸 Challenge 3: Spotify Authentication

 Issue: Spotify’s OAuth2 setup required proper redirect URI even for testing.

 Solution: Used a developer token temporarily and documented the OAuth setup for future
deployment.

🔸 Challenge 4: Docker Port Exposure

 Issue: Docker container wasn't accessible externally after EC2 deployment.

 Solution: Opened required ports in EC2 security groups and exposed Streamlit’s port
properly in the Dockerfile.

5.6 Code Documentation and Versioning

The project code was managed with Git and GitHub, ensuring every update was version-
controlled. Highlights:

 Proper README.md file created.

 Used .env files to manage API keys securely.

24
  Branch-based development for major modules (voice, maps, music).

 All modules had inline comments and docstrings for clarity.

 Final build was zipped and pushed to GitHub for long-term access and sharing.

5.7 Project Evaluation & Feedback

During the final review at Regex Software Services:

 My code was found to be clean, modular, and scalable.

 Mentors appreciated the effective use of Streamlit and Docker, technologies not
commonly mastered at the student level.

 Suggestions were made to add emotion detection and user profiling in future versions.

 The presentation was well received, with an emphasis on how multi-API integration was
handled effectively.

5.8 Final Reflection

This industrial training was not just about building a chatbot — it was a journey of professional
growth. It helped me:

 Work independently on complex technical problems.

 Follow structured software engineering practices.

 Handle voice, language, maps, music, and cloud — all in one project.

 Gain the confidence to now develop, test, and deploy complete AI applications.

25
REFERENCES

1. Regex Software Services – Official Website

2. Google Generative AI API Documentation

3. Spotify for Developers – API Docs

4. Streamlit Python Framework

5. Speech Recognition Library Documentation

6. Folium Mapping Library

7. Geopy Geocoding Library

8. Docker Official Documentation

9. AWS EC2 User Guide

10. Python Packages: pyttsx3, dotenv, openai, gtts, spotipy

26