Ages 18-25
Anxious about studies, early career choices and life changes.
ACADEMIC TEAM PROJECT · TELEGRAM
PRODUCT DISCOVERY / EMOTIONAL SUPPORT
Anna✳
Making the first step toward emotional support easier while keeping the AI assistant’s boundaries clear.
AI prototypeCustomer journeyFeedback loops
a.Anna · AI companiononline now•••
CONVERSATION PROTOTYPE
Hi. What has been taking up most of your thoughts lately?
I feel tired of rushing all the time.
It sounds like you have a lot to carry. Where would you like to start?
conceptual conversation visualWrite a message…↗
01 / PROBLEM & AUDIENCE01
Support may be needed now, while reaching out can feel hard.
The team explored a digital support concept for people facing cost, time or personal barriers to in-person counselling. The concept is a conversational Telegram bot for reflection and simple self-help exercises.
Three audience hypotheses
Busy professionals
Experience stress but struggle to fit regular sessions into their schedule.
Cautious explorers
Want to understand their feelings privately and are not ready to speak with a specialist.
02 / CUSTOMER JOURNEY
Trust is built step by step.
The journey maps describe two different paths: actively looking for quick support and cautiously exploring help in private.
Notice
A person notices stress or exhaustion and looks for a low-pressure way to reflect.
Compare
They assess privacy, tone, credibility and whether they can try it without a commitment.
Start
A short introduction and clear explanation of the bot’s role lower the first-message barrier.
Use
A supportive dialogue, exercises and retained context may make it easier to return.
Decide
A sense of safety shapes whether they continue or recommend the service.
PERSONA AThe Overwhelmed AchieverPERSONA BThe Cautious Explorer
03 / RESEARCH SYNTHESIS03
A calm tone is a start. Trust needs more.
Low-friction entry
People need a simple start, clear expectations and a chance to try a few conversations before paying.
Privacy & control
The cautious persona is especially sensitive to confidentiality, a non-judgmental tone and transparent data handling.
Dialogue, not canned replies
Feedback asked for clarifying questions, a more active bot and continuity across conversations.
AI boundaries
Participants question whether AI can replace human empathy. The product’s role must be stated honestly.
04 / PROTOTYPE
A supportive dialogue inside a familiar messenger.
In the concept, a user opens a Telegram bot, learns what it can do and talks through stress or anxiety. The bot asks reflective questions and suggests small exercises. The project materials list the OpenAI API.
TELEGRAM×OPENAI API
Welcome
Who the bot is, what it can help with and where its limits are.
Conversation
The user describes a situation at their own pace.
Reflection
A follow-up question helps the user express the feeling more clearly.
Exercise
A short self-help practice as an optional next step.
05 / FEEDBACK → ITERATION05
Separate shipped changes from the next hypotheses.
Feedback notes mention a polite tone, a simple interface and comfortable conversations. Users also asked to delay the subscription offer, let the bot ask questions, add personalization and retain context.
Changes listed in the project report
Defined three segment hypotheses in place of one broad audience.
Clarified the proposition: accessible emotional support and self-help skills in Telegram.
Reworked the copy, added reflective exercises and made the bot’s capabilities easier to understand.
Added a clear statement that the bot may support users but does not replace professional therapy.
Possible next iterations from feedback
Delay the subscription prompt and let people explore the flow first.
Explore ways to adjust the bot persona, such as its name and form of address.
Test follow-up questions and gentle mood check-ins.
Explore retaining useful context without collecting excessive sensitive data.
!
PROJECT BOUNDARIES
This is an academic AI prototype for emotional support and self-reflection, not a medical or diagnostic service. It does not replace therapy or professional care. The case presents a concept and user feedback; effectiveness and safety were not clinically evaluated.
SOURCES: JOURNEY MAPS, DESIGN THINKING FEEDBACK AND TEAM PRESENTATION