Paul Ostergaard, CEO, Norwood Systems, 30 April 2025.
As I speak with carriers from Singapore to Toronto, one refrain is constant: “We know we need AI – how do we operationalise it fast enough?”
GSMA Intelligence estimates that 81 percent of Tier‑1 operators now have an AI roadmap, yet barely 15 percent have scaled beyond pilot projects (prod-cms.gsmaintelligence.com).
Meanwhile, the rest of industry is moving at break‑neck speed; McKinsey’s 2024 global survey shows generative‑AI adoption across all industries leapt from 35 percent to 65 percent in just ten months (McKinsey & Company). Subscribers are already enjoying AI‑driven experiences from tech giants, and they now expect the same frictionless intelligence from their network provider.
The CSP Dilemma: Adapt or Risk Irrelevance
OTT competition is no longer an early warning on the horizon; it is here, scaling at internet speed. TeleGeography’s International Voice Report counted ≈ 6 billion monthly users across just seven OTT comms apps (WhatsApp, Messenger, WeChat, QQ, Viber, Line, KakaoTalk) as of September 2024 – and that total excludes FaceTime, Google Meet and Skype (TeleGeography’s Official Blog).
For operators, the financial drag is visible in the hard numbers:
-
Mobile voice revenue is projected to slide from US $244 billion in 2025 to US $211 billion by 2029, a –3.6 % CAGR, according to Statista’s 1Q‑25 market update (Statista).
-
PwC’s Global Telecoms Outlook 2024‑28 expects sector‑wide ARPU to keep shrinking ~2 % a year until 2028, even as total line counts rise – a clear sign that value is leaking up the stack to OTT and cloud players (PwC).
In short, CSPs that stick to legacy voice economics are on a one‑way path to commoditisation while Apple, Google and Meta capture the customer relationship – and the margin – through AI‑powered services delivered at software speed. The clock is ticking.
The Norwood Advantage: Accelerate Your AI-Driven Transformation
Norwood Systems designed OpenSpan to turn “AI ambition” into revenue‑ready reality!
Built natively on existing hyperscaler infrastructure, OpenSpan drops a modular, carrier‑grade AI layer (voice, messaging, analytics) straight into an operator’s existing OSS/BSS stack. That means no forklift swaps: a micro‑services SIP-IMS gateway securely proxies SIP, WebRTC or 5G traffic into AI pipelines for transcription, fraud scoring, real‑time translation and intent orchestration (TelecomTalk).
Where operators would traditionally open a twelve‑to‑eighteen‑month waterfall programme, OpenSpan ships with pre‑built agents, including call transcription, SME answering services, call‑fraud firewalls to name a few. As a result, CSP teams can move from “sandbox” to commercial launch in a few quarters.
The platform’s commercial levers are equally pragmatic. Usage‑based billing APIs let a finance team price services per minute, per call, per month or per fraud check – whichever aligns with local market norms – while carrier‑grade multi‑tenant controls ensure regulators, auditors and enterprise customers see the same security posture they expect from core voice switches. Our early adopter modelling shows a projected 30-45 % incremental gross margin uplift on AI‑bundled voice products versus legacy tariffs, primarily driven by reduced churn and premium feature‑tier ARPU.
“OpenSpan exemplifies how CSPs can leverage Microsoft Azure’s powerful AI and cloud capabilities to redefine their voice service offerings. By working with Norwood Systems, we are enabling CSPs to accelerate innovation, deliver personalized customer experiences, and transform the way they operate in a rapidly evolving digital landscape.”
Rick Lievano, CTO, Worldwide Telecommunications Industry, Microsoft
Real Revenue Opportunities for CSPs
Consider a few examples of immediate AI-driven services that CSPs can deploy rapidly using OpenSpan, each driving tangible revenue uplift, customer retention, and churn reduction:
1. Advanced Call Screening and Fraud Prevention
Telecom fraud is no longer a niche irritant – it is a $39 billion‑a‑year drag on the industry, with losses jumping 12 percent in just two years (AI Receptionist Service). At the retail edge, the volume is staggering: Hiya flagged 11.3 billion suspicious calls in Q4 2024 alone, and that is only the traffic it could see (Hiya).
OpenSpan lets an operator move the defensive line from the handset to the network core. Real‑time AI models score every inbound and outbound call in‑flight, blocking spoofed CLI, shutting down Wangiri rings, and triggering step‑up authentication when a threshold is breached. The immediate payoff is a measurable fall in complaints and involuntary churn – fraud is one of the top three reasons subscribers port away.
Just as important, the operator can re‑monetise trust by offering a “clean‑line” tier to high‑value subscribers and enterprises: a small per‑subscriber fee quickly compounds into a multi‑million‑dollar annual service with minimal marginal cost.
2. AI-Enhanced Customer Interaction Platforms
Nearly half of consumers now expect round‑the‑clock support, and they punish brands that cannot deliver it (see contact-centres.com). At the same time, boards remain laser‑focused on cost: McKinsey finds cost control still ranks in the top three priorities for customer‑care leaders in 2024 (McKinsey & Company).
OpenSpan’s white‑label SME Assistant squarely addresses both pressures. A single API call spins up an always‑on voice agent that greets callers by name, triages intent, books appointments into the SME’s calendar, and pushes transcripts into the CRM – all in the customer’s preferred language.
Because the assistant is hosted in the operator’s cloud tenancy, usage minutes and data stay on‑net, turning what used to be low‑margin connectivity into sticky platform revenue. For the SME, it wipes out the cost of out‑of‑hours staff; for the operator, it unlocks an ARR‑style licence plus incremental voice traffic, delivering healthy margin with minimal acquisition spend.
3. Dynamic, Personalised Voice Services
Generic IVRs and prerecorded clips no longer cut it. Customers now expect the same hyper‑personalisation in voice that they get in digital apps – and they are willing to pay for it. Microsoft’s January 2025 analysis shows that operators offering contextual voice assistants see a direct uplift in average revenue per user (ARPU) as subscribers opt into premium feature bundles (Microsoft).
OpenSpan’s real‑time generative engine taps live intent, profile data, and third‑party APIs to create truly conversational journeys: a travel customer can rebook a flight mid‑call, an elderly patient can confirm a clinic slot without navigating a website, and a utilities customer can renegotiate a plan based on real‑time usage – all without human intervention. Each transaction produces incremental billable events (minutes, API hits, data queries) and – more importantly – cements the operator’s position at the centre of the customer’s digital life, driving loyalty and reducing churn over the long term.
The Urgency of Immediate Action
Standing still is no longer an option. Operators that hesitate risk surrendering customer intimacy – and the accompanying revenue – to faster, more agile rivals. By deploying OpenSpan, a CSP can compress innovation cycles from years to weeks, push AI‑powered services to market ahead of OTT competitors, and recapture the narrative of who truly owns the customer relationship.
CSPs that embed AI‑rich voice and messaging today will out‑innovate OTT players, unlock fresh revenue streams, and cement unbreakable customer loyalty.
Your Next Move
If you are serious about protecting – and expanding – your share of the digital wallet, the conversation starts now. Our global telco team is ready to map OpenSpan’s capabilities to your network roadmap, demonstrate quick‑win use cases, and craft a deployment plan that delivers measurable ROI within months.
Contact your Norwood account executive today and let’s accelerate together into telecom’s AI‑defined future.
— Paul Ostergaard, CEO, Norwood Systems