HSE Chatbot: faster access to schedules and student services
Designing a conversational interface that helps students find academic information and use digital services without navigating multiple sections.
Project overview
HSE App is HSE University's mobile app for students and instructors. It already includes core academic services such as schedules, search, profiles, and other tools supporting the educational experience.
Context
Students regularly need quick access to academic and administrative information: checking a schedule, finding the right service, or resolving an academic question without a lengthy search.
This information is often spread across different areas of the app, the university website, and the student portal. Even a simple request can require several steps and take more time than expected.
This is especially important on mobile, where students need direct access to answers without navigating the full service structure.
Challenge
Design a chatbot for HSE App that helps students find academic information and access digital services faster.
The solution needed to be more than another chat screen. It had to serve as a practical assistant for everyday student tasks, helping users resolve academic requests and access core services without unnecessary navigation between sections.
What I did
- Analyzed common academic and administrative student requests;
- Defined the chatbot's core scenarios;
- Mapped the customer journey for schedule lookup;
- Designed user flows for schedules, office hours, and digital services;
- Defined the tone of voice and messages for fallback and error states;
- Designed the chatbot interface and its core interaction scenarios.
Problem
1. Information was fragmented across services
Students had to retrieve academic information from several sources. Class schedules lived in one place, document requests in another, instructor and classroom information somewhere else, and some content was available only on the website.
As a result, students spent time finding the right entry point instead of completing the task itself.
2. Mobile tasks required too much effort
Some tasks could be completed through the website or student portal, but those paths were not always convenient on a phone. When a page was poorly adapted to a mobile screen, students had to zoom, locate the right option, and complete a long flow in an awkward format.
This created a gap in HSE App: the app was always within reach, but it did not always support quick academic tasks from end to end.
3. Students did not always know where to look
Even when a feature existed, users might not know its official name or where it was located. Students thought in natural requests rather than interface categories: “an instructor's schedule for Friday,” “how do I request a certificate?” or “when is the dean's office open?”
The interface therefore needed a way to understand user intent and guide students from a natural-language request to the right result.
Research
Personas
I developed three key personas representing different HSE App use cases. The personas helped keep the design focused not only on product functionality, but also on users' real needs, motivations, and pain points.
Customer journey map
I created a customer journey map to understand the student's path through the app and chatbot. This helped identify key touchpoints, uncover potential barriers, and find opportunities for the interface to guide users to the right result faster.
Design
User stories
- As an HSE student, I want to check my schedule quickly so that I can plan my time effectively.
- As an HSE student, I want to check classroom availability so that I can find a room for additional classes.
- As an HSE student, I want to check instructor schedules so that I can attend additional classes.
- As an HSE student, I want to check other groups' schedules quickly so that I can attend additional classes with my cohort.
- As an HSE student, I want to request the document I need through the mobile app so that I can receive it sooner.
- As an HSE student, I want to check the dean's office hours quickly so that I do not waste time searching.
Tone of voice
Because the chatbot serves an educational product, its tone needed to be respectful, clear, and approachable. An overly formal style would make the assistant feel cold, while too much humor could feel inappropriate for academic and administrative requests.
I chose a balanced voice: calm, polite, human, and sparing with humor. The bot addresses students respectfully, communicates outcomes quickly, and avoids unnecessary detail.
User flow: My schedule
The user selects a quick action or types a message. The bot checks the date and displays the class schedule. If no data is available, it explains that the schedule has not been published yet or that there are no classes that day.
User flow: Schedule lookup
This flow supports searches for an instructor, group, or classroom schedule. The user enters the details, the bot asks a clarification question when needed, and then displays the schedule for the selected date.
User flow: Office hours
Users can type a natural question such as “When is the dean's office open?” or select a suggested prompt. The bot finds the relevant department and displays its current office hours.
User flow: Digital services
A user can enter a request such as “I need proof of enrollment.” The bot finds the relevant service and offers to submit the request online or explains the next step.
Edge cases
I also defined fallback and error states for the chatbot:
- No internet connection;
- Voice request not understood;
- Voice input temporarily unavailable;
- Service maintenance;
- Information not found.
This matters because users judge an assistant by more than successful responses. When the bot cannot complete an action, it still needs to explain what happened and offer a clear next step.
Solution
Chatbot as a new search layer
The chatbot replaces the conventional search page and expands what search can do. Instead of locating the right section, students can ask a question in their own words and receive a direct answer.
The opening screen surfaces popular scenarios as quick actions, helping users start common tasks without typing while immediately communicating what the assistant can do.
Natural-language queries
In addition to quick actions, the chatbot supports short, natural-language queries. Students do not need to phrase a complete question or know the exact name of a section. They can simply type “tomorrow's schedule,” “dean's office hours,” or “proof of enrollment.”
When a request lacks enough detail, the bot asks a clarification question instead of immediately showing an error. For example, it may ask the user to choose from several instructors with similar names or specify a date.
This approach aligns the interface with familiar search behavior and reduces unnecessary steps.
Voice input
Voice input supports moments when typing is inconvenient: while walking, moving between campus buildings, waiting for class, or commuting. Users can speak a request, and the bot processes it like regular text.
Light and dark themes
The conversational interface includes multiple visual states, including light and dark themes. This keeps the chatbot consistent with app settings and makes the conversation comfortable across different environments.
Outcome
The project delivered a chatbot for HSE App that helps students complete academic and administrative tasks faster.
The core value of the solution is a simpler, more natural way to access university services. Search no longer behaves like a conventional list of sections and pages, but as an assistant that understands user intent, helps clarify a request, and guides students directly to the right result.