AI‑Powered Conversational Assistant for Admission Automation in

Government Educational Platforms

N. Ganitha Aarthi

, Thasni Asharif

, Shyam Subbiah

, Shifa Ashwath

Sitheshwaran

and Narendra Nath

Department of Computer Science and Design, SNS College of Technology, Coimbatore, Tamil Nadu, India

Department of Computer Science, Dr. SNS. Rajalakshmi College of Arts and Science, Coimbatore, Tamil Nadu, India

Department of Computer Science and Design, SNS College of Engineering, Coimbatore, Tamil Nadu, India

Keywords: Government Administration, Rule‑Based Inferencing, Assistant System, Regulatory Information,

Multilingual Support, Decision‑Making Efficiency, Public Sector Automation.

Abstract: The Admission Guru project is an AI-powered assistant integrated into the Rajasthan Education Department’s

website to simplify admission-related processes for engineering and polytechnic institutes. It provides instant,

accurate responses to queries on eligibility, college options, fees, scholarships, and placements using advanced

NLP and AI technologies. Accessible 24/7, the multilingual assistant supports English and Hindi, ensuring

inclusivity for diverse users. By automating responses to frequently asked questions, it reduces administrative

workload and enhances efficiency. Admission Guru streamlines the admission process, improves user

satisfaction, and serves as a transformative tool in Rajasthan’s educational ecosystem.

1 INTRODUCTION

In government administrative departments, officers

often encounter queries requiring reference to

specific rule books and their interpretations. This can

be a time-consuming and error-prone process,

especially when the information is scattered across

various documents and languages. Our proposed

solution aims to streamline this process by developing

a Assistant system that assists officers in providing

accurate and contextually relevant information based

on government rules and regulations. The feedback

for this system comes from real-time interactions with

officers and administrators who frequently deal with

regulatory queries. They have expressed that

answering these queries often relies on their

experience and familiarity with the rules, which can

be challenging and time-consuming. Various rules

and regulations must be consulted, and interpreting

them accurately is crucial to ensure compliance and

informed decision-making. Officers have reported

that the current process is tedious and can lead to

delays in administrative functions.

2 OBJECTIVES

Students and parents in Rajasthan face difficulties in

accessing accurate, up-to-date information about

departmental vacancies, eligibility criteria, and

scholarship opportunities. The existing system is

fragmented, complex, and hard to navigate, leading to

confusion, delays, and missed opportunities. A

solution is needed that simplifies the admission

process, centralizes all relevant information, and is

accessible and user- friendly for all users.

3 LITERATURE REVIEW

The rise of digital solutions has transformed

education, with chatbots emerging as efficient tools

for automating processes, addressing inquiries, and

enhancing accessibility. In the context of student

admissions, the Admission Guru chatbots leverages

natural language processing (NLP) and artificial

intelligence (AI) to deliver real-time, multilingual

support for students and parents. Chatbots utilize

NLP to interpret user inputs and generate relevant

responses, as detailed by Jurafsky and Martin (2019)

in Speech and Language Processing. While NLP

170

Aarthi, N. G., Asharif, T., Subbiah, S., Ashwath, S., Sitheshwaran, and Nath, N.

AI-Powered Conversational Assistant for Admission Automation in Government Educational Platforms.

DOI: 10.5220/0013909700004919

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT‘25 2025) - Volume 4, pages

170-175

ISBN: 978-989-758-777-1

facilitates language comprehension, implementing it

for Indian languages like Hindi introduces challenges

such as tokenization, syntax parsing, and semantic

analysis (Sharma & Verma, 2021). These challenges

are critical for creating inclusive, multilingual

systems.

Shah and Singh (2020) underscore the role of

chatbots in streamlining administrative workflows

and providing round-the-clock support for admission-

related queries. Tools like the Google Gemini API

enable context- aware interactions and real-time data

integration, essential for systems like Admission

Guru. Frameworks like Streamlit further enhance user

engagement by offering responsive, interactive

interfaces (Kaushik & Gupta, 2022). The scalability

and reliability of chatbots are essential, particularly

during admission peak periods. Cloud-based

infrastructures ensure high traffic handling without

compromising performance (Kaushik & Gupta,

2022). Voice- enabled features and predictive

analytics improve accessibility and personalize user

experiences by anticipating frequently asked

questions (Shah & Singh, 2020; Sharma, 2021).

Security and privacy are vital for protecting

sensitive user data. Russell and Norvig (2021)

advocate encryption protocols and ethical AI

practices to ensure user trust. Analytics also play a

crucial role in identifying performance gaps, refining

chatbot capabilities, and improving user engagement

(Kaushik & Gupta, 2022).

4 METHODOLOGY

4.1 Existing System

The admission process as shown in figure 1 begins

with Step 1, where applicants are required to attend

the Parent Orientation and secure the application

form. In Step 2, applicants must submit all the

Admissions Requirements to the Admissions Office.

At this stage, admission personnel will contact the

applicant to schedule the entrance test. Moving to

Step 3, applicants need to pay the processing fee at

the cashier and complete the Entrance Test. After

three (3) working days, applicants must contact the

Admissions Office to obtain their test results, as

outlined in Step 4. If successful, the admissions

personnel will call the applicant to schedule an

interview in Step 5. Finally, in Step 6, applicants

receive their Acceptance Letter and proceed to

enrolment.

Figure 1: Existing System Block Diagram.

This process involves approximately six steps,

including two interactions (test and interview) and a

three-day waiting period for test results. By following

these steps, applicants efficiently complete the

admission process.

Pros of Existing System: The structured admission

process offers clarity and transparency, with clear

expectations at each step. Parent orientation ensures

families are informed, while the entrance test and

interview maintain academic standards by admitting

qualified candidates. The defined timeline, including

a three-day waiting period, helps both applicants and

the admissions office prepare for subsequent steps.

Cons of Existing System: The admission process

involves delays, multiple interactions, and waiting

periods, causing anxiety for applicants. The three-day

wait for test results adds uncertainty. Applicants must

manage several steps independently, which can be

overwhelming. Additionally, the process lacks

flexibility for those requiring special

accommodations, reducing accessibility for diverse

applicants.

AI-Powered Conversational Assistant for Admission Automation in Government Educational Platforms

171

4.2 Proposed System

4.2.1 User Interface of Prototype

Currently, the Admission Guru chatbot is hosted on

Streamlit Community Cloud (figure 2), a cloud based

platform that enables easy deployment and access to

web applications. This deployment choice makes the

chatbot publicly accessible to users via a provided

link, allowing anyone with an internet connection to

interact with the assistant. The use of Streamlit

Community Cloud eliminates the need for users to

install any software locally, simplifying the process

of accessing the chatbot. Users can directly access the

chatbot through a browser, providing a seamless

experience for anyone looking for admission-related

information, no matter where they are located.

Figure 2: Streamlit Cloud App Ui.

The cloud-based deployment also ensures that the

chatbot is available around the clock. Since the

chatbot is hosted in the cloud, there is no dependency

on local systems or infrastructure. This means that

users can engage with the bot at any time, whether it's

during regular office hours or outside of them. Figure

3 shows the Admission Guru UI in Website.

Figure 3: Admission Guru UI in Website.

When a user clicks the chat icon, a popup window

appears with a simple, intuitive chat interface. The

layout is minimalistic and responsive, keeping the

user focused on the conversation. At the bottom, a

text input field allows users to type their queries,

which can be submitted via the "Enter" key or a

button. The chatbot responds instantly, displaying

answers in a clear, conversational format. The

messages are shown chronologically for easy

understanding. Visual elements, such as avatars or

speech bubbles, may be included to make the

interaction feel more natural and engaging for users.

Figure 4 gives the Data-Flow Diagram.

Figure 4: Data-Flow Diagram.

Upon clicking the icon, a popup window will appear,

displaying a simple and user-friendly chat interface

where users can type their queries and view the

assistant’s responses in a conversational format. After

submitting a query, the chatbot will instantly display

relevant answers retrieved in real-time from live

website data and multiple PDF documents, processed

via the Gemini API. If a query cannot be answered, it

will provide a fallback response, encouraging the user

to refine their question or try another query.

Implementation Stack is shown in figure 5.

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

172

Figure 5: Implementation Stack.

It allows for rapid development of interactive web

applications with minimal effort. Streamlet’s

simplicity makes it a perfect choice for building

intuitive, user-friendly interfaces that don’t require

extensive frontend development expertise. For the

Admission Guru assistant, Streamlit is used to create

the chat interface where users can input their queries

and receive responses from the AI- powered assistant.

The application is embedded into the website, and the

chat interface allows for seamless, real-time

conversations. Streamlit also supports the addition of

multilingual capabilities, ensuring that users can

interact with the assistant in both Hindi and English,

with potential for more languages in the future. The

platform’s flexibility allows the chatbot to be

scalable, handling varying user traffic during peak

periods, such as the admission season. Figure 6 shows

the system architecture block diagram.At the core of

the architecture is the AI Processing Layer, powered

by the Google Gemini API. This layer handles natural

language processing (NLP) tasks, including intent

recognition, context understanding, and response

generation. It ensures that user queries, whether

simple or complex, are accurately interpreted and

addressed. 29 The AI Processing Layer interacts with

a structured Data Management System, which

currently comprises PDF documents containing

frequently asked questions and answers related to

admissions. In the future, this layer will integrate with

live data from the Department of Technical

Education, Rajasthan's website to provide real-time

updates and insights.

Figure 6: System Architecture Block Diagram.

5 RESULTS

In terms of the methods or algorithms used to solve

the problem, ChatGPT-3.5 implements the rational

roots theorem five out of six times and Cardano’s

formula once. ChatGPT-4 attempts to provide a

solution by using the rational roots theorem, a

graphical solution, and a code snippet in python

66.7%, 16.7%, and 16.7% of the time, respectively.

Finally, Gemini AI uses factor lists five times and the

rational roots theorem once. All the implemented

methods or algorithms can correctly lead to a right

answer; thus, it could be said that the chatbots have

chosen a proper way to give an answer. Figure 7

shows the Comparison Chart.

AI-Powered Conversational Assistant for Admission Automation in Government Educational Platforms

173

Figure 7: Comparison Chart.

6 CONCLUSIONS & FUTURE

ENHANCEMENT

6.1 Conclusions

The Admission Guru project, developed as an inbuilt

admission assistant for the Rajasthan Education

Department's website, has successfully fulfilled its

objective of streamlining the admission process for

students and parents. By integrating features such as

viewing department vacancies, scholarship details,

and admission guidelines, we have significantly

enhanced user experience. The design thinking

process allowed us to empathize with the target

audience, define their pain points, ideate solutions,

prototype the assistant interface, and test its

functionalities to ensure it meets the needs of

students, parents, and educational administrators.

Our approach began by understanding the

challenges faced by users in navigating the traditional

admission process. These challenges included lack of

timely information, complexity in finding relevant

department vacancies, and confusion regarding

scholarship opportunities. By directly addressing

these issues, Admission Guru was designed to

provide real-time, easily accessible information in a

user-friendly interface. The prototype was validated

with a set of test users, receiving positive feedback

regarding its simplicity and efficiency in guiding

them through the admission journey.

We also incorporated continuous feedback loops,

using real-world data to refine the system's

functionalities and ensure that the platform remains

intuitive and effective. As a result, students and

parents are now able to access critical admission

details seamlessly, making the overall process more

transparent and less stressful. Educational

administrators benefit from the tool as well, since it

automates many manual processes, reducing the

workload and enhancing operational efficiency.

However, while Admission Guru has successfully

met its initial objectives, there is always room for

improvement. The foundation laid in this project can

be further expanded upon to integrate more advanced

features and improve the user experience even more.

6.2 Future Enhancement

The future of Admission Guru holds exciting

possibilities. There are several enhancements that

could elevate the tool to new levels of usefulness for

both users and administrators.

Multilingual Support: As Rajasthan is a diverse

state with a multitude of languages spoken, adding

multilingual support could make Admission Guru

more accessible to a wider audience. This feature

would enable users from different linguistic

backgrounds to interact with the platform in their

preferred language. Integration with Other Systems:

Admission Guru could be integrated with other

education management systems to provide a more

comprehensive service. For example, linking the

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

174

assistant to databases that track students' academic

performance and other eligibility criteria could

automate the admission process even further.

Mobile Application: While the tool is accessible via

the website, a dedicated mobile application could be

developed to allow users to easily access the platform

on the go. Given the high mobile penetration in

Rajasthan, this could increase the tool’s adoption

among students and parents.

AI-Powered Guidance: Incorporating AI to provide

personalized advice could enhance the user

experience. The assistant could analyze a student’s

preferences, academic history, and eligibility to

recommend specific departments, courses, and

scholarships.

User Feedback Mechanism: Adding a robust user

feedback mechanism would allow continuous

improvement. Users could report issues or suggest

new features, helping the development team prioritize

future updates based on real user needs.

Enhanced Analytics for Administrators: The

platform could provide educational administrators

with more detailed insights and analytics regarding

admissions, such as trends in application volume,

preferred departments, and common queries. This

could help streamline decision-making and resource

allocation.

REFERENCES

Gupta, S., & Kumar, A. (2020). "Cloud Infrastructure for

Scalable Chatbot Applications." International Journal

of Cloud Computing, 14(1), 75-90.

Jurafsky, D., & Martin, J. H. (2019). Speech and Language

Processing. Pearson.

Kaushik, M., & Gupta, A. (2022). "Optimizing Admission

Processes through Digital Chatbots: A Case Study."

International Conference on Digital Transformation in

Education.

Russell, S., & Norvig, P. (2021). Artificial Intelligence: A

Modern Approach. Pearson.

Shah, P., & Singh, K. (2020). "AI-Powered Chatbots in

Education: Enhancing User Engagement and

Accessibility." International Journal of Educational

Technology, 12(4), 215-230.

Shah, P., & Patel, M. (2021). "Voice- Enabled Chatbots:

Bridging the Digital Literacy Gap." Journal of Human-

Computer Interaction, 33(7), 501-520.

Sharma, A., & Verma, R. (2020). "Enhancing

Chatbot Security and Privacy for EducationalApplicati

ons." Proceedings of the IEEE Symposium on Secure

AI Systems.

Sharma, A., & Verma, R. (2021). "Challenges in

Multilingual NLP: A Focus on Indian Regional

Languages." Journal of Artificial Intelligence Research,

45(2), 120-135.

Sharma, A. (2021). "Multilingual Conversational AI

Models for Education: Opportunities and Challenges."

AI in Education Review, 9(3), 132-150.

Verma, R., & Singh, V. (2022). "Real-Time Data

Integration in Chatbots: Techniques and Challenges."

IEEE Transactions on AI Systems, 18(6), 450-462.

AI-Powered Conversational Assistant for Admission Automation in Government Educational Platforms

175