Agentforce Deployment for Enterprise Service Operations
National Insurance Provider
Business Context
The organization and its strategic environment
A $9B insurance provider handling 2.4M customer service interactions annually needed to reduce call center costs while improving resolution quality. Traditional chatbots had failed — customers hated them. Salesforce Agentforce offered a fundamentally different approach: AI agents that actually resolve issues, not just deflect them.
Technology Landscape
Systems and infrastructure before DGT
Salesforce Service Cloud (existing), 3 legacy claims systems, IVR platform, knowledge base with 12,000 articles, and a workforce of 1,800 service agents across 4 contact centers.
The Challenge
What the client was facing
Previous chatbot deployments achieved only 12% containment (88% of interactions still required human agents). Customer satisfaction with digital channels was 2.1/5. The company needed AI that could actually resolve complex insurance queries — policy changes, claims status, coverage questions — not just route them.
The DGT Solution
How DGT addressed the challenge
DGT deployed Salesforce Agentforce with deep integration into claims systems, policy administration, and knowledge base. We built 14 specialized AI agent 'skills' covering the top 80% of service interactions. Each skill was trained on 3 years of resolved case data and validated by senior agents before deployment.
DGT Accelerators Used
Delivery Approach
How DGT executed the engagement
'Agent-Led AI' methodology: we partnered human agents with AI agents during a 6-week pilot. Human agents coached the AI, flagged errors, and validated responses. This created a feedback loop that improved AI accuracy from 71% to 94% before full deployment.
Governance Model
How the engagement was managed
AI Ethics Board oversight, weekly accuracy reviews, human-in-the-loop escalation for complex cases, and continuous learning pipeline with agent feedback integration.
Timeline & Phases
The execution roadmap
AI Readiness Assessment
3 weeksData quality audit, use case prioritization, skill mapping
Skill Development
8 weeks14 Agentforce skills built, trained on 3 years of case data
Human-AI Pilot
6 weeks200 agents paired with AI, feedback loop, accuracy improvement
Phased Rollout
8 weeksContact center by contact center deployment
Optimization
OngoingContinuous learning, new skill development, accuracy monitoring
Risks Addressed
Key risks DGT mitigated during the engagement
Why DGT Won
What set DGT apart in this engagement
Outcome Metrics
Measurable before-and-after results
| Metric | Before | After DGT |
|---|---|---|
| AI Containment Rate | 12% (chatbot) | 67% (Agentforce) |
| Customer Satisfaction | 2.1/5 | 4.3/5 |
| Average Handle Time | 8.2 min | 3.1 min |
| Annual Service Cost | $52M | $34M |
The Impact
Headline results delivered
"The difference between our old chatbot and Agentforce is night and day. Customers actually prefer the AI agent for routine queries now — it's faster, more accurate, and available 24/7. DGT made this possible."
Patricia Hernandez
SVP Customer Operations, National Insurance Provider