Customer support · Production

AI Agent Ida handles 28% of SpareBank 1 Østlandet's customer inquiries with 41% automation rate

The problem

In 2018, SpareBank 1 Østlandet faced growing customer demand for digital service channels and competitive pressure as peer banks deployed chatbots, while the bank itself had only implemented robotic process automation and lacked a conversational AI capability.

Workflow diagram · grounded in source
1
Customer submits inquiry
trigger
“Customers don't want to search through endless FAQs anymore. They want to ask a question and get an answer. Ida makes that possible”
2
Ida handles known questions
ai_action
“the AI Agent answers questions across banking, insurance, and savings”
3
Generative AI for unknown intents
ai_action
“plays a key role in managing unknown intents, questions that fall outside the model's predefined scope, and provides contextual, relevant responses using generative AI”
4
Guardrail compliance validation
validation
“responses are generated within pre-approved guardrails that prioritize both customer safety and regulatory compliance”
5
Human advisor escalation
routing
“can seamlessly connect customers to human advisors when needed”
6
Generative AI transparency labeling
output
“all conversations with the AI Agent that are powered by generative AI are clearly labeled so customers know they are interacting with a large language model”
Reported outcome

Ida became the bank's second-largest customer service channel, handling 28% of all customer inquiries with a 41% automation rate and over 2.3 million conversations since launch, delivering service at a fraction of the cost of traditional channels and reducing manual workload for AI trainers.

Reported metrics
customer inquiries handled by Ida28%
Automation rate41%
Total conversations since launchover 2.3 million
conversations since December 1 202485,000
Show all 9 reported metrics
customer inquiries handled by Ida28%
automation rate41%
total conversations since launchover 2.3 million
conversations since December 1 202485,000
conversations powered by generative AI10%
total generative AI conversationsover 6,800
cost of service delivery vs traditional channelsfraction of the cost of traditional channels
telephony inquiry share40%
manual AI trainer workload reductionreduced manual workload for AI Trainers
Reported stack
boost.aiGenerative ActionIda
Source
https://www.boost.ai/case-studies/how-sparebank-1-ostlandet-is-writing-the-playbook-for-generative-ai-adoption-in-banks
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Ida became the bank's second-largest customer service channel, handling 28% of all customer inquiries with a 41% automation rate and over 2.3 million conversations since launch, delivering service at a fraction of the…

What tools did this team use?

boost.ai, Generative Action, Ida.

What results were reported?

customer inquiries handled by Ida: 28%; Automation rate: 41%; Total conversations since launch: over 2.3 million; conversations since December 1 2024: 85,000 (source-reported, not independently verified).

How is this customer support AI workflow structured?

Customer submits inquiry → Ida handles known questions → Generative AI for unknown intents → Guardrail compliance validation → Human advisor escalation → Generative AI transparency labeling.

Related customer support cases