← Sticker Wall AGENT · DETAIL

Campus Psychological Center AI Night Shift & Crisis Referral System: A B2B Subscription Generating 10K RMB/Month

Workflow: Every night, the system automatically takes over the student venting portal on the psychological center's WeChat Officia

AGENT

Key Fields

FIELD STAMPS

IndustryHealthcare / Elderly Care

RegionChina（中国大陆）

ScaleSME

ChannelOnline

🔧 Workflow

Every night, the system automatically takes over the student venting portal on the psychological center's WeChat Official Account or Mini Program: the AI first scans for crisis keywords and classifies emotions into green, yellow, and red status levels. Green triggers automated empathetic companionship and writes an emotional diary; yellow enters the next day's follow-up queue; red instantly generates a referral work order and pushes it to the on-duty counselor via WeChat Work. After the counselor handles the case, they write back the results in the work order system. Every week, the AI automatically compiles a risk report and data dashboard for the head of the psychological center.

🛠 Setup Requirements

Requires basic backend development skills or proficiency with workflow tools like Coze and n8n; apply for a WeChat Service Account and configure dialogue capabilities; first, train a crisis keyword library tailored for scenarios like postgraduate exam failure or romantic breakups, then integrate with Large Model APIs for deep emotional classification, and achieve real-time work order pushing via the WeChat Work API. The MVP version takes about 1-2 weeks to complete basic listening and tiering functions, followed by 2-4 weeks of closed-beta piloting with a university psychological center to fine-tune dialogue scripts and referral workflows.

🧰 Toolchain

🔧 Coze
🔧 WeChat Official Accounts Platform
🔧 DeepSeek API
🔧 n8n
🔧 WeChat Work

💰 Revenue

Targeting a starting monthly revenue of 10,000 RMB, with a subscription fee of about 3,000-5,000 RMB/year per university psychological center. Securing 3-5 contracts achieves stable income; in the early stage, a free pilot can be offered to a university in exchange for a long-term subscription contract based on data reports. The AI emotional companion side hustle has already proven that the emotional companionship track can generate over 10,000 RMB/month (youres.cn), while the university B2B model features higher average order values and more stable renewal rates.

💸 Cost

Estimated monthly fixed costs under 500 RMB: Large Model API calls around 100-200 RMB (varying by consultation volume), lightweight cloud server around 80 RMB/month, WeChat verification fee of 300 RMB/year amortized to 25 RMB/month; costs can be compressed to under 200 RMB/month during the initial pilot phase.

⏱ Time Investment

Investing about 2-3 hours daily: reviewing high-risk work orders from the previous night and AI false negatives in the morning; handling the follow-up queue and updating the dialogue script library at noon; monitoring system operational status and training the morning emotional report for the next day in the evening; spending 50 minutes on Sunday compiling the weekly risk trend report for the psychological center.

🚀 Getting Started

Step 1 for beginners: do not write code yet. Spend 3 days researching nighttime emotional support scenarios in target universities, find the head of the psychological center or teachers responsible for crisis intervention, and trade a free AI duty tool for a pilot opportunity. Step 2: Quickly build a venting bot on the Coze platform featuring risk tiering and referral forms, run it as a gray-scale release in student communities for a week, and focus on recording high-risk keywords missed by the AI as well as misclassified normal venting.

🔑 Keys to Success

✅ Closed-loop crisis tiering and human review: it is better to over-refer than to miss a single high-risk case; a human safety net mechanism must be in place
✅ Penetrate using the trust endorsement of the psychological center or counselors to resolve data privacy sensitivity and ethical responsibility boundaries
✅ Continuously update high-frequency emotional keywords and stage-specific scripts for postgraduate exam failures and breakups to avoid the backlash triggered by generic AI responses
✅ Establish strict anonymization and data encryption processes to comply with the ethical standards and regulatory requirements of university psychological counseling

⚠️ 风险

⚠️ Crisis intervention involves personal safety and privacy compliance, and misjudgment can lead to severe consequences. A disclaimer agreement must be signed and 27/4 human online review capability must be maintained
⚠️ University psychological centers have long procurement processes and complex decision-making chains; going from pilot to contract may take 3-6 months, putting pressure on cash flow
⚠️ If users perceive AI empathy as mechanical responses, it can lead to a collapse of trust. Persona settings, memory retention, and emotional copywriting design are needed to reduce the robotic feel
⚠️ If the system crashes or API rate limits occur during night duty, a backup plan involving human hotlines and SMS must be prepared to prevent emergency calls from going unanswered

📌 Real Cases

📌 36Kr Report: A leading emotional companionship app has accumulated 60 million users and hundreds of millions in revenue, validating the scalable payment potential and market acceptance of the AI emotional companionship track
📌 TRAE Competition Project 'Xinyu': A 24-hour AI mental health companionship website for college students, deployed in pilot tests across multiple universities, proving that real and unmet demand exists for campus emotional companionship scenarios
📌 Dukou AI Project: Built a breakup healing and companionship system based on the Kübler-Ross grief curve, guiding users from the acute phase to the reconstruction phase through a stage-adaptive dialogue model, validating the feasibility of AI healing in niche emotional scenarios
📌 youres.cn Report: AI emotional companion has become a side hustle generating over 10,000 RMB per month, proving the commercial viability of individuals and micro-teams entering the market via vertical emotional companionship

Sources