GSOC ‘24 Proposal: City of

Boston

Table of Contents
Table of Contents
Section 1: Information About Me
Section 2: My Civic Commitment
My passion for civic issues
Civic Technology in My Community
My Exposure To Civic Technology
Section 3: My Project Skills
Idea 1: Requesting City Services via AI-Driven Apps
Technical Architecture for AI-powered 311 App
Skills and Technologies associated
Technical Skills
Non-Technical Skills
My concluding remark

GSOC ‘24 Proposal: City of Boston 1

 Section 1: Information About Me
Name: Arunanshu Sinha
Country: India
LinkedIn: https://www.linkedin.com/in/arunanshu-sinha/
Contact Information

Email Id: [email protected]

Mobile Phone no: +91 9689621868

Resume: https://drive.google.com/file/d/1Y5H__PH3Q5oF4LzHdlblUMfK1bUSn-
Fk/view
Skills:

Language / Library / Framework Experience Level (Out of 5)

Python 3.5

SQL 3

Numpy 3.5

Pandas 3.5

Matplotlib 3.5

Seaborn 3

Scikit-learn 3.5

TensorFlow 3

PyTorch 2.5

Keras 2

Django 3

Flask 2

Github: https://github.com/sirius-arun
Academic Institution: Bachelors of Technology, Indian Institute of Technology
(BHU)
Work Experience:

GSOC ‘24 Proposal: City of Boston 2

 Product Management Internship at Languify AI (Building conversational AI for
education)

Technical Product Management Internship at Monetizely Inc. (B2B SaaS

Startup disrupting pricing and CPQ)

Strategy Consulting Internship at WeGoT Technologies ( Building Smart water

metering solutions)

Section 2: My Civic Commitment

My passion for civic issues
From a young age, I've been fascinated by the potential of a strong society, where
opportunity thrives and everyone has a chance to contribute their unique talents.
This passion has led me to three interconnected civic issues that I hold immense
intent to solve for

Education

Infrastructure

Inclusivity & representation.

Education is the cornerstone. Quality education empowers individuals and

strengthens communities. However, disparities in access and resources can leave
many behind. I believe in fighting for equitable education, ensuring every child has
a safe and well-equipped learning environment, regardless of background. This
includes advocating for infrastructure projects that bridge the digital divide and
provide equal access to technology, a crucial tool for modern learning.

Infrastructure is the foundation. Strong infrastructure connects us, fosters

economic opportunity, and allows communities to flourish. But when infrastructure
crumbles or fails to meet the needs of all residents, it creates barriers. I'm
passionate about ensuring infrastructure projects are inclusive, considering the
needs of diverse communities and promoting sustainable practices for a healthy
future.
Inclusivity & representation are the heart of it all. A society that reflects the
richness of its people is a society that thrives. I believe in promoting diversity in
leadership roles, both in education and infrastructure development. Seeing role

GSOC ‘24 Proposal: City of Boston 3

 models who look like them can ignite a spark in young minds, inspiring them to
pursue careers that shape their communities.
These issues may seem separate, but they are deeply intertwined. Strong
infrastructure supports quality education, and both are meaningless without
inclusivity. Working for a government organization that tackles these
interconnected challenges is incredibly exciting to me. I see it as an opportunity
to be a part of something bigger, to contribute to building a society where
potential is unlocked, not limited. My passion for these issues fuels my
dedication.

Civic Technology in My Community

I am a resident of a City called Pune in Maharashtra, India. Below are some basic
stats:

Pune has been previously tagged as

Land area 331.26 sq.km
the “Oxford of East”, and is the 9th
Population 6.99 million (2022)
most populous city of India. Pune has
Languages Marathi, Hindi, English also been labelled as the most liveable
GDP US$ 69 billion (2021) city in India recently.

As a resident of this city that is actively adopting technology to improve the lives
of it’s community members I have witnessed the power that civic technology
holds.

As a resident, I have witnessed that Pune's civic bodies are already transitioning
to technology, and I know that there's immense potential to leverage it for a more
efficient and citizen-centric experience. Here's how I see civic technology
improving life in our community:

Hyperlocal issue reporting: Develop a mobile app for citizens to report and
track community issues like potholes, overflowing bins, or damaged
streetlights. Integrate GPS and image capture to ensure swift action by
relevant authorities. This empowers citizens and fosters transparency.

Data-driven decision-making: Implement data dashboards for real-time

monitoring of civic services like water supply, traffic flow, and public transport.

GSOC ‘24 Proposal: City of Boston 4

 This allows authorities to identify problem areas and allocate resources
strategically, leading to optimized service delivery.

AI-powered grievance redressal: Utilize AI chatbots to address frequently

asked questions and guide citizens through the grievance redressal process.
This frees up human resources for complex issues and provides 24/7 support,
improving citizen satisfaction.

Participatory budgeting platform: Create an online platform for citizens to

propose and vote on local development projects. This fosters inclusivity and
empowers residents to shape their community's future directly.

Multilingual accessibility: Ensure all civic technology solutions are accessible

in Marathi and other prevalent local languages. This bridges the digital divide
and empowers a wider range of residents to participate.

My passion lies in using technology to solve real-world problems. These are just
a few examples, and I'm eager to learn from the innovative approaches used by
the US government. I believe combining my technological expertise with your
organisation's experience can significantly improve lives in Pune and beyond.

My Exposure To Civic Technology

My first real-life exposure to seeing civic technology being developed was 2 years
ago when my father was working with Maharashtra’s Chief Minister Dash Board
Project. It was an interactive dashboard through which the chief minister’s office
could minutely review any work right up to the district level. The ultimate aim is to
make the dashboard an interface between the citizens and the government,
making the information and statistics on the dashboard open to the public.

Throughout the development of this project, both of us brainstormed on multiple

use cases and their feasibility and ways of implementation. This sparked an
interest in civic technology in me.

Link to the report: https://theprint.in/politics/fadnavis-gets-a-digital-dashboard-

to-keep-eye-on-projects-ahead-of-maharashtra-polls/273084/

GSOC ‘24 Proposal: City of Boston 5

 Section 3: My Project Skills
Though both the ideas in the list are extremely intriguing, my personal interest
gravitates more towards Idea 1.
My Excitement

This idea excites me because it has the potential to significantly improve the
efficiency of Boston's 311 system. By streamlining the reporting process, we can
empower residents to voice concerns quickly and ensure city services are
deployed effectively.

My Suitability
As someone passionate about civic technology, I'm fascinated by the possibility of
using AI to bridge the gap between residents and city services. My background in
technical product development and deep interest in machine learning makes me
confident in tackling the technical challenges of building an accurate and user-
friendly system.
My Motivation

Having personally experienced the frustration of using an inefficient system, I

believe this idea can make a real difference in people's lives. Simplifying the
reporting process can encourage more civic engagement and improve the overall
quality of life in Boston.

My Vision
This project aligns perfectly with the growing trend of AI-powered citizen
services. I'm eager to be a part of developing a solution that could be a model for
other cities looking to leverage similar technology for the public good.
My Intention

This project is technically fascinating to me for a couple of reasons. Firstly, it

allows me to apply my recent deep dive into Convolutional Neural Networks
(CNNs) and their image-processing capabilities. CNNs have revolutionized image
recognition tasks, and I'm eager to put this knowledge to work by training a model
to identify specific service requests in Boston's 311 system. Secondly, the

GSOC ‘24 Proposal: City of Boston 6

 challenge of optimizing the model for real-world deployment on a mobile app is
particularly interesting. It requires considering factors like efficiency, accuracy,
and scalability, which aligns perfectly with my interest in practical AI applications.

Idea 1: Requesting City Services via AI-Driven Apps

Technical Architecture for AI-powered 311 App

This project involves building an image recognition system with a user-friendly
interface for Boston's 311 app. Here's a breakdown of the technical architecture:

The Components:

1. Mobile/Web App Frontend:

Users can capture or upload images.

Interface to interact with the image recognition API.

Displays potential service categories based on image analysis.

Allows users to submit requests with relevant details.

2. Image Preprocessing Pipeline:

Handles image uploads from the app.

Performs operations like resizing, cropping, and normalization.

Converts images to a format suitable for the machine learning model.

3. Image Recognition API:

Core component using a trained machine learning model.

Receives preprocessed images from the pipeline.

Analyzes the image and predicts the most likely service category.

Outputs a ranked list of potential service categories.

4. Machine Learning Model:

Trained on a dataset of past 311 service requests with corresponding

images.

Uses image recognition techniques (explained below) to classify images.

GSOC ‘24 Proposal: City of Boston 7

 5. Backend Server:

Hosts the image preprocessing pipeline and image recognition API.

Communicates with the mobile/web app frontend.

Might connect to existing 311 system database for service details.

Building the Machine Learning Model:

Convolutional Neural Networks (CNNs): Most suitable for image recognition

tasks.

Other Popular options: ResNet, VGG, and Inception.

Python Libraries and Frameworks:

Image Preprocessing: OpenCV, Pillow (PIL Fork)

Machine Learning Backend: TensorFlow, PyTorch

API Development: Flask, FastAPI

Development Breakdown:

1. Data Collection and Preprocessing:

Gather past 311 service requests with corresponding images (ensure data
privacy regulations are met).

Label images with appropriate service categories.

Preprocess images for model training (resizing, normalization, etc.).

2. Machine Learning Model Training:

Choose a pre-trained CNN model and fine-tune it on the prepared dataset.

Train/validate/test the model to ensure good accuracy and performance.

3. API Development:

Develop an API endpoint using Flask/FastAPI to receive images and return

service category predictions.

Integrate the trained model for making predictions within the API.

GSOC ‘24 Proposal: City of Boston 8

 4. Mobile/Web App Development:

Build the user interface for capturing/uploading images and interacting

with the API.

Display predicted service categories and allow users to submit requests.

5. Deployment and Integration:

Deploy the backend server (API + preprocessing pipeline) to a suitable

platform (local server or cloud).

Integrate the mobile/web app with the deployed API for communication.

Skills and Technologies associated

Technical Skills Non-Technical Skills
Python Programming Gritty Problem solving

Machine Learning Effective Communication

Deep Learning Frameworks Commitment & Integrity

API Development Prioritisation

Deployment Creativity & thinking outside the

box
Some additional knowledge of
performance monitoring tools, data Presentation & writing
labelling tools, version control
systems

I am in my 3rd year of pursuing engineering and have been a part of over 4 teams
developing technical products via internships and hackathons mostly in the
capacity of a Product manager. Including my current role at a seed-funded deep-
tech AI startup working on building world-class conversational AI for education.
I may not already have a great level of expertise in the technical skills mentioned
above but my work experience over the past couple of years has inculcated a
great work ethic in me! A testament to that is this recommendation note by my
previous employer who is also a Staford Grad and the founder of Monetizely Inc. (I

GSOC ‘24 Proposal: City of Boston 9

 worked with him as a technical product manager at the company for 9 months
remotely)

My concluding remark
The Google Summer of Code programme has been uplifting the open source
development culture for 20 years. I am extremely willing to work with City of
Boston via GSoC 2024. Thank you for taking the time to assess my proposal. I
hope I can add value and work on the idea proposed.

GSOC ‘24 Proposal: City of Boston 10