Get in Touch
← Back to case studies

Customer Service · Agentforce case study

Connecting website AI chat to Salesforce service cases

Customers ask the website Agentforce chatbot about orders and service issues. Supported questions receive guidance; unresolved requests become Salesforce cases for the service team to continue.

  • Salesforce
  • Agentforce
  • Service Cloud
  • Human handoff

At a glance

Business need
Help website visitors while giving unresolved requests a human follow-up path
Salesforce scope
Agentforce chatbot and Salesforce case handling
Connected systems
Website chat, approved service information and Salesforce CRM cases
Delivery focus
Chatbot guidance, case escalation and service-team follow-up

The situation

A website question should have a clear next step

A customer opens the Agentforce chatbot on the company website to ask about an order or a service issue. The AI provides an answer or guidance when it can. If the question cannot be handled by the chatbot, it moves into a Salesforce case for the customer service team to continue.

Website chat beside a Salesforce-style case composer and conversation summary.
Illustrative Salesforce-style interface with synthetic sample content. It depicts the described workflow, not the actual client system or a standard product screen.

The challenge

An unanswered chat must reach the service team

When the chatbot cannot resolve an order or service question, the customer needs a route to further help. The implementation connects that moment to a Salesforce case so the service team can take the request forward.

The solution

Agentforce connects website help to Salesforce case follow-up

Shinetech connected the website Agentforce chatbot to the Salesforce case process. Customers use the chat for order and service questions. Supported questions receive an answer or guidance; unresolved requests become cases for the customer service team to continue.

Service staff also use AI in their work. The customer-facing chat and the service case provide two connected ways to handle the request, with a person responsible for follow-up when the chatbot cannot resolve it.

How it works

01 · Ask

The customer asks the website chatbot about an order or service issue.

02 · Answer

The AI gives an answer or next step for a supported question.

03 · Escalate

An unresolved question becomes a Salesforce case.

04 · Follow up

A service agent continues the request from the case.

How we work

Design the customer conversation and the service handoff together

The delivery approach starts with the order and service questions customers bring to the website. The service team identifies which questions the chatbot should handle and which need a person. Approved information and response boundaries guide that design.

Review follows both customer paths: useful guidance for a supported question, and a Salesforce case for a request requiring service follow-up. The check is whether each path gives the customer an appropriate next step and the service team a request it can continue.

The result

Website questions have a route into customer service

Customer

Can ask order and service questions through the website and receive guidance or a route to further help.

Service team

Receives unresolved requests as Salesforce cases for follow-up. Staff also use AI while handling customer requests.

The chatbot becomes an entry point to the service process, with case-based follow-up for questions that need a person.

Case taxonomy

Explore the industry, technology, and business need.

Industry and product

  • Cross-Industry
  • E-commerce

Technology and delivery

  • Salesforce Agentforce
  • Service Cloud
  • AI Chat

Business need

  • AI Customer Service
  • Approved Answers
  • Human Handoff
  • Case Follow-up