🌍 Context

HouseBox is a home goods e-commerce company receiving around 900 customer inquiries per month through email, Telegram, and phone calls.

All requests were handled manually: support agents had to read each message, determine its urgency, and decide who should respond. The support team consisted of three agents, but growing workloads led to employee burnout, and one team member eventually left the company.

The business already used Gmail, Slack, and Google Sheets, but the customer support process remained entirely manual.

⚠️ Problem

As the number of customer inquiries increased, the support team struggled to keep up with demand.

Key challenges included:

• Average response times had reached 11 hours.
• Approximately 50% of inquiries were repetitive “Where is my order?” requests.
• Urgent customer issues were often buried among lower-priority messages.
• Management only became aware of problems after customers submitted complaints.
• Customer satisfaction was declining, creating reputational risks for the business.

The company owner was considering hiring two additional support agents, but budget constraints made this option difficult.

💡 Solution

I designed and implemented an AI-powered customer support workflow that automatically processes incoming inquiries and ensures timely follow-up.

The system analyzes each new request, identifies its category and priority, assigns it to the appropriate team member, and automatically applies a response deadline based on predefined service levels.

If a ticket is not addressed on time, the system escalates the issue through multiple notification levels, ensuring visibility and accountability across the support team.

By automating ticket triage and prioritization, the support team no longer spends time sorting and routing requests manually. Instead, agents can focus on solving customer problems and handling more complex cases.

Priority Levels & SLA

🔴 P1 — 30 minutes