Customer Service Automation
analyze thousands of long customer-service calls at reasonable cost so the business could understand intent, sentiment, and service pain points at scale.
Reported impact
Jet2 deployed its first LLM use case in three months, analyzes about 2,000 calls a day for intent, and found a lower-cost way to maintain high-quality sentiment classification at scale.
Jet2 uses Snowflake and Cortex AI on Snowflake to analyze thousands of long customer-service calls at reasonable cost so the business could understand intent, sentiment, and service pain points at scale. Thousands of daily call transcripts from Jet2 contact centers are stored and analyzed in Snowflake to reveal customer intent, preferences, and pain points. Cortex AI supports the core summarization and classification workflow, and Jet2 maintains an LLMOps framework to tune performance and hold down cost as prompts and classifications evolve. Analysts and service leaders use the outputs to improve customer service, and the same framework is being extended to other communications and self-service insight use cases across the company. Keeping ingestion, storage, transformation, and governed access on one platform reduces handoffs, improves concurrency, and makes it easier to reuse trusted data across teams and applications. Jet2 deployed its first LLM use case in three months,...
contact-center call transcripts flow into Snowflake and Cortex AI; curated data and model outputs are then served to analysts, service leaders, and future messaging-channel workflows.
2,000 Calls analyzed daily for customer intent
I don’t think we could have gone down the LLM route at all without Cortex
With no prior experience in LLMs, Jet2 was able to deploy its use case in just three months using Cortex AI
using Cortex AI, Jet2 has been making the most of large language models to analyze customer intent and sentiment across thousands of call transcripts
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
OpenAI Sample Modified · Oct 16, 2026 Localiza Localiza needed to turn generative AI into a company-wide operating model that could improve internal work, create new conversational customer entry points, and support AI-native products without treating the technology… 130 squads ChatGPT · Codex · Liza OpenAI Sample Oct 8, 2026 Virgin Atlantic At Virgin Atlantic, we’ve always seen ourselves as a challenger—willing to take bold bets that help us stand apart. AI fits that philosophy perfectly. ChatGPT Enterprise · Codex · Realtime API 
ServiceNow Sample Modified · Aug 28, 2026 Southeastern Railway Southeastern Railway was processing Delay Repay claims, complaints, incident reports, and fraud checks across disconnected systems, which delayed compensation, forced temporary hiring during disruptions, and made trend… 70% of Delay Repay claims automated Customer Service Management · ServiceNow Otto · Virtual Agent 
