Leading Travel Organization
Knowledge retrieval for a leading travel organization
Summary
Instant, case-aware information on demand, reducing agent search time, cutting repeat calls, and delivering max €700K estimated ROI.
The customer support human agents at this travel organization spent significant time searching for the right information during live calls, leading to longer handling times, inconsistent answers, and repeat customer calls. DEUS designed an AI-enabled experience that detects customer intent in real time, auto-surfaces relevant content through context filters and prompt suggestions, and presents clear, actionable answers, with a built-in feedback loop for continuous quality improvement.
Services
Industry
The Challenge
Human agents losing valuable call time searching for answers, leading to inconsistency, repeat contacts, and higher handling costs.
Human customer support agents faced a structural productivity challenge: during live customer calls, they had to manually search multiple knowledge sources for relevant information. This created delays, increased average handling time (AHT), introduced inconsistencies in the answers given, and resulted in repeat contacts from customers who didn't get a complete answer the first time.
Problem 01
High search overhead during live calls
- ▪Agents interrupted call flow to manually search knowledge bases
- ▪Inconsistent retrieval paths across team members
- ▪Increased average handling time across all contact types
- ▪Agent cognitive load reduces quality of customer interaction
Problem 02
Inconsistent answers driving repeat contacts
- ▪Different agents retrieving different answers to the same question
- ▪Customers calling back for corrections or clarifications
- ▪Repeat contacts inflating contact center volume
- ▪No systematic feedback mechanism to improve knowledge quality
The Solution
DEUS designed an AI-enabled knowledge retrieval experience that works in real time during customer calls, detecting intent, surfacing relevant content, and presenting actionable answers directly to the agent without interrupting the conversation flow.
01
Real-time intent detection
The system listens to the live conversation context and automatically detects what the customer is calling about, classifying intent before the agent has to search manually.
02
Context filters and knowledge retrieval
Based on detected intent, the system applies context filters and retrieves the most relevant information from the knowledge base, surfacing prompt suggestions and actionable answers directly in the agent interface.
03
Feedback loop for continuous improvement
Agents can rate the relevance and accuracy of retrieved content, feeding a continuous improvement loop that increases knowledge freshness and retrieval precision over time.
Impact
Estimated €700K max ROI by reducing search overhead and repeat contacts, while improving the quality and consistency of customer interactions.
Search Time
Estimated value of reduced agent search time, driving lower average handling time (AHT) across all contacts.
Repeat Calls
Estimated value of repeat call reduction at 20% industry average, consistent, accurate answers on first contact.
Diagnostic KPIs
Conservative to optimistic improvement scenario, tracked across 8 second-order diagnostic KPIs including accuracy, freshness, and latency.
Methods
A human-centered, end-to-end design and delivery approach, grounding every decision in agent reality and organizational process.
User interview
Shadowing
Focus group
Personas
Journey map
Service blueprint
Stakeholders map
Ideation workshop
Concept book
Design sprint
UI / UX design
ROI modeling
Solution architecture
Engineering
Project management