Industry

E-Commerce/Retail Tech

Client Location

United Kingdom

Project Duration

2.5 Months

Engagement Model

Time and Material

Team Size

3 AI Specialists, 1 UX Designer, 1 QA Engineer

Client Background:

The client is a UK-based online retailer of lifestyle, home, and fashion goods. The brand had to deal with an increase in customer assistance inquiries across all channels. Due to its expanding client base and frequent product introductions, particularly during the holiday and sales seasons. Their small customer support crew was overworked. Which resulted in disgruntled consumers, delayed responses, and higher operating expenses.

The Vision:

The goal of the brand's leadership was to use an AI chatbot connected to their website. Website to automate up to 70% of consumer inquiries. Without sacrificing the caliber of customer service the objective was to improve the user experience. Clients offer round-the-clock assistance and lessen the manual labor on the support staff.

Project Brief:

The business required a more sophisticated, multilingual AI chatbot that could comprehend complex queries, access order records, and make product recommendations. They had earlier tried with simple chat widgets. Their main goals for the MVP were as follows:

• Order tracking and status updates that are automated
• FAQ automation (problems with refunds, shipping, and payments)
• Astute product suggestion
• Escalation of difficult issues to human actors
• Zendesk with Shopify integration
• Before a full-scale launch, they asked us to create and produce an MVP in eight weeks to test its viability.

Development Challenges:

• Teaching the chatbot product-specific terms and regional slang.
• Ensuring a seamless transition to human operators while avoiding consumer annoyance.
• Connecting to a variety of data sources, including logistics APIs, CRMs, and Shopify stores.
• Creating a chatbot user experience that reflects the tone and voice of the brand.
• Meeting deadlines while adhering to quality assurance procedures

Solutions Provided:

To identify the most frequently asked questions. Our team worked closely with the client's e-commerce and support teams. What we included in the MVP is as follows:

• AI Chatbot Framework: Developed with Google Dialogflow CX and a Node.js backend to gather data dynamically.
• Previous NLP Model Training: specially educated using chatbot dialogues and data from previous help tickets.
• Multi-Channel Deployment: Using Twilio and Meta APIs, the website and social media messengers are integrated.
• CRM and Order Sync: Instantaneous data retrieval and ticket changes with real-time connectivity with Zendesk and Shopify.
• The analytics dashboard allowed the client to monitor escalation rates, dropped chats, and chatbot performance.



Key Results:

• Within the first month of the MVP launch, the number of daily support tickets was reduced by 60%.
• Agents can concentrate on high-value tasks because 92% of order status inquiries are handled automatically.
• For popular requests, response times are three times faster.
• Consistent round-the-clock assistance across all channels leads to increased customer satisfaction.