Normalized claim
Client engagement uplift: 2 x increase
Discovery AI doubled client engagement with Discovery Bank next best actions.
MicrosoftEvidence: Medium50/100
Discovery Bank: Discovery AI hyper-personalization with Azure OpenAI in Foundry Models
Use case typeCustomer personalizationUpdated Jun 13, 2026
Discovery Bank needed to scale hyper-personalized financial experiences and deliver faster, smarter client interactions without managing complex infrastructure. Using Azure OpenAI in Foundry Models and Azure Databricks, Discovery Bank built Discovery AI, a generative AI application that powers personalized recommendations for clients and helps service agents tailor their interactions with customers. Discovery AI doubled client engagement with Discovery Bank next best actions. The AI-powered experience reduced latency of response times by over 50% and improved client satisfaction through real-time, personalized financial insights.
- Organization
- Discovery Bank
- Industry
- Finance
- Location
- South Africa
- Published
- June 2026
Reported outcomes
2x
client engagement upliftCustomer experience
−50%response latency reduction3,000 questions/daydaily questions processed+70%client recommendation engagement2xtraffic growth
Strategic outcomes
Customer experience & trustImproved personalized client interactionsSpeed & agilityReduced response latencyNew product / capabilityBuilt AI-powered personalized recommendations
Why do we believe this?Outcome claims, sources, and evidence checks
Normalized claim
Response latency reduction: 50% decrease
The AI-powered experience reduced latency of response times by over 50%.
Normalized claim
Daily questions processed: 3,000 questions/day
Discovery Bank’s servicing agents now process roughly 3,000 questions daily.
Normalized claim
Client recommendation engagement: 70% increase
70% of clients engaging with personalized recommendations.
Normalized claim
Traffic growth: 2 x increase
traffic through Discovery AI nearly doubled in the past month.
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- Discovery Bank
- Provider
- Microsoft
- Maturity
- Unknown
- Linked source
- Microsoft Customer Stories
No explicit deployment-stage evidence found.
Customer identity supportedSource describes one deploymentMaturity evidence evaluated
Primary read
Use case focus
Showing 3 of 4
- 1Customer Service
- 2Personalization
- 3Agent Assist
- Scale hyper-personalized financial experiences and deliver faster, smarter client interactions without heavy infrastructure management.
- Enable contact-center agents to provide personalized responses and next-best actions.
- Built Discovery AI using Azure OpenAI in Foundry Models and Azure Databricks.
- Combined behavioral modeling with generative AI to generate personalized recommendations and agent-assist responses.
- Used client profiles and a knowledge base to tailor responses in the Discovery Bank app and via WhatsApp 24/7.
- Fine-tuned multiple Azure OpenAI 4o-mini and 4.1-mini models to reduce latency and improve domain accuracy.
- Used multimodal inputs such as photos, voice notes, and documents for tailored responses.
- Doubled client engagement with next best actions.
- Reduced response latency by over 50%, from about 5–6 seconds to about 1.5–2 seconds.
- Agents process roughly 3,000 questions daily.
- About 70% of clients engage with personalized recommendations.
- Traffic through Discovery AI nearly doubled in the past month.
Architecture
Discovery Bank runs Discovery AI on Azure Container Apps and Azure App Service, with Azure API Management as the secure front door for AI requests. The solution connects contact-center workflows, Azure Databricks, and Azure OpenAI endpoints. Azure Cosmos DB stores unstructured chat data and ongoing conversation context, Azure Data Factory handles streaming and ETL between on-premises systems and the cloud, and Azure Key Vault protects credentials and secrets. Azure AI Document Intelligence is planned for KYC use cases.
Sources & evidence1
Evidence: Medium50/100Evidence strength
- Customer explicitly identified
- Primary source available
- Quantified outcome available
- Technical implementation details available
Type: Customer StoryPublished: Jun 7, 2026Publisher: MicrosoftEvidence: PrimaryConfidence: High
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Customer evidence
Provider evidence
AI patterns
Browse the catalog
Was this useful?
Community
Comments
Loading comments...