{"slug":"building-trust-in-mobile-communications","url":"https://findausecase.com/use-cases/building-trust-in-mobile-communications","title":"Building trust in mobile communications","description":"First Orion, which protects mobile subscribers from scam calls by analyzing over 350 million calls daily, migrated from AWS EMR to the Databricks Data + AI Platform to power anomaly detection, fraud prevention and branded-calling products, boosting productivity by more than 35% and generating over $12.7 million in value over three years. The company is also using open LLMs such as MosaicML MPT and Dolly on the lakehouse to build conversational AI for document search and scam interception, with preliminary results showing a 5x increase in search efficiency.","company":"First Orion","industry":"Technology & Software","aiCapabilities":["Fraud & Anomaly Detection","Recommendation & Personalization","Large Language Models","Conversational AI"],"businessFunctions":["Marketing","Sales"],"technology":["Databricks Data + AI Platform","Delta Lake","Databricks SQL","MLflow","Unity Catalog","Agent Bricks","MosaicML MPT","Dolly"],"deployment":"Public Cloud","problemStatement":"As data volumes increased, First Orion's legacy reliance on AWS EMR and other native services became overwhelming to manage, with data silos, duplicate data and complex infrastructure making scaling expensive and slow. Clusters took one to six hours to provision, jobs queued and productivity stalled, and clusters ran 24/7 to meet peak workload demands, leading to over-provisioned compute costing as much as an average of $6,000 per user per month. Granting secure access under telecom compliance standards and ISO certification was a full-time job, and anonymizing certain roles in data restores exploded the size of their data sets and cost to manage them.","solutionApproach":"First Orion migrated to the Databricks Data + AI Platform, gaining automated cluster management, support for multiple programming languages, shared collaborative notebook environments, and easier access controls and governance. Over 200 employees across more than eight IT, marketing and sales teams now run analytical reports, train ML models, automate billing and build hundreds of ETL pipelines on the platform, using Delta Lake as the storage layer, Databricks SQL for BI workloads, MLflow for the ML lifecycle, and Unity Catalog for governance, access control, auditing and lineage. Its INFORM and ENGAGE branded communication solutions rely on anomaly detection, monitoring, fraud prevention and brand personalization, while SENTRY uses a call protection API integrated with Databricks for call analytics and billing to spoof-proof outbound numbers. The team is also using open models such as MosaicML MPT and Dolly to build LLMs that replace manual document search and target AI answering and scam interception.","businessValue":"Branded communication built on the lakehouse architecture raised answer rates by 200%, and INFORM increased first-call conversion rates by 76%. First Orion has boosted productivity by more than 35%, and operational efficiencies and new product innovations have contributed more than $12.7 million in value over three years. Preliminary results from LLM-based document search show a 5x increase in search efficiency.","evidence":{"band":"high"},"sourceUrl":"https://www.databricks.com/customers/first-orion","dates":{"publishedAt":"2026-09-22T05:48:08.427Z","publishedAtSource":"pipeline","updatedAt":"2026-09-22T05:48:08.427Z"},"license":"Open for reading and citing with a link to https://findausecase.com/use-cases/building-trust-in-mobile-communications. Bulk republication requires permission."}