Solutions
Alepo Digital BSS with AI-Powered CX
Telco Type
Leading ICT provider (digital MVNO brand)
Country
Canada
Launching a Digital-Only Mobile Brand: SaskTel’s Lüm Mobile
A case study of how SaskTel, a leading ICT provider in Saskatchewan, Canada, launched Lüm Mobile, an all-digital, membership-based mobile service with no stores, no agents, and no monthly billing dates, and then pushed the model further with a generative AI customer assistant. Prepared for operators evaluating a digital sub-brand or AI-led customer service.
91% lower
OPEX than traditional wireless services
70%
of service volume handled by AI within two months
25% lower
churn after the AI assistant deployment
+10%
monthly subscriber growth in the first quarter
Background
SaskTel saw growing demand for self-service, more affordable plans, and environmentally conscious options, and a younger, digitally native segment its traditional brand was not built to serve. Rather than retrofit the main brand, it launched a separate digital-only brand: no stores, no human agents, all interactions through a website and app with automated support.
The Business Model Decision
The distinctive choice was not the technology. It was the model. Four decisions shaped the program:
Membership, not plans
No monthly plans, no billing dates, no long-term contracts. Subscribers buy a membership, then buy data only when they need it, and it never expires. A model with structurally higher margins than traditional wireless.
Zero-touch as the cost structure
With sales, service, and support delivered entirely through digital channels, the cost base of stores and call centers is removed by design, not optimized after the fact. OPEX runs 91% below SaskTel’s traditional wireless services.
BYOD and eSIM by default
Bring-your-own-device and eSIM support reduce device logistics and electronic waste, aligned with the brand’s environmental positioning.
Managed operations
End-to-end management of the IT systems and private cloud sits with the platform partner, keeping the operator’s in-house teams focused on the business rather than infrastructure.
Program Scope
The service runs on a digital BSS deployed within the operator’s private cloud: billing and charging, product catalog, policy control, CRM, brand website, web and mobile self-care, order and inventory management, payment integration, promotions and loyalty, community forum, analytics, and managed services, integrated with the MNO core and IT systems as a single platform. Subscribers self-manage the full lifecycle: onboarding, number portability, BYOD and eSIM activation, top-ups, and support.
Launch Results
Within three months of launch, the service was growing subscribers by 10% each month, reaching digitally native segments, including students and young people, the traditional brand had not captured, with customer care and IT support running at a fraction of regular operations. The launch generated significant organic social media attention, lifting brand visibility and engagement.
AI Customer Assistant
In a service with no stores and no live agents, support quality is the brand. In January 2024, Lüm Mobile replaced its first-generation, rule-based chatbot with Lümbot, a generative AI customer assistant built on a vertically tuned large language model, integrated directly with the backend IT systems so it can complete tasks on a customer’s behalf, not just answer questions. Fine-tuning and integration brought error rates below human-agent levels while increasing task completion, resolving most issues at first contact, with fewer escalations to human agents.
Within two months of deployment, the AI assistant was handling 70% of customer service volume and contributing to 30% of overall sales. 89% of interactions received positive or neutral sentiment ratings. Churn fell 25%, and AI-driven upsells added a 20% increase in revenue.
“With the introduction of Lümbot, we are taking a big leap forward in our digital journey, employing generative AI and LLM technology to enable a holistic, end-to-end digital customer service experience.”
Charlene Gavel, President and CEO, SaskTel
The 91% OPEX gap came from stores and call centers never being built: the digital brand is a cost structure, not just a channel.
The assistant’s value came from integration: wired into the backend systems, it completes tasks rather than deflecting them.
Error rate and task completion were benchmarked against human agents before the AI went live on the service.
What to Verify When Selecting a Digital BSS and AI Assistant
Checks drawn from this program, useful regardless of vendor:
Will the platform let the brand run as a new cost structure, or just put an app on the old one?
Can your AI assistant complete tasks in backend systems, or only answer questions?
What share of service volume can be resolved at first digital contact, measured rather than assumed?
How are AI error rates benchmarked against your human agents before go-live?
Does the AI contribute to revenue (upsell, retention), or only to cost reduction?
