Netcall Plc · LSE:NET
Netcall — CX Live 2026
13 slides from Netcall's conference presentation, Jul 21, 2026
Slide 1: From Paralysis to Progress
From Paralysis to No silver bullet. Progress No magic dust. Just better operating discipline. What’s really standing between your contact centre and AI that delivers results? Richard Farrell | Chief Innovation Officer | Netcall © Netcall
Slide 2: AI didn’t start with ChatGPT
AI didn’t start with ChatGPT ASR & TTS NLU IBM Watson NLP Bots LCAP & AI ML Transformers LLMs Agentic AI Late 1990s Mid 2000s Mid 2010s Mid 2010s Late 2010s Early 2020s Early 2020s 2022 Mid 2020s Netcall has been successfully embedding AI into enterprise solutions for more than 20 years © Netcall
Slide 3: The technology has evolved. The fundamentals have not.
The technology has evolved. The fundamentals have not. Each wave changes what AI can do but value still comes from connecting technology to real work, controls and outcomes. What has evolved What has not changed Why this matters now • Recognition • Start with known demand • The room already knows sp
Slide 4: The room has heard the AI story
The room has heard the AI story So this session has to be more useful than another prediction about the future. What the show says What leaders need Netcall’s experience • Agentic AI • Where to start • Start with known demand • Predictive engagement • What to measure • Connect AI to real work • Cont
Slide 5: AI needs an operating model, not just a model
AI needs an operating model, not just a model The simple, proven pattern behind successful use cases. Pick known demand Connect to the work Prove the outcome Start where the volume, Give AI context, a bounded Measure resolution, effort, pain, and baseline are action path, and the right quality, cost
Slide 6: How not to prove AI value
How not to prove AI value Why AI pilots need harder evidence than self-reported minutes saved What the pilot claims Why the method is flawed What would prove value • Starts with the technology • Named operational problem 43 • Uses self-reported time savings • Pre/post baseline • Measures gross minut
Slide 7: Outcome example: Rotherham NHS DNA AI pilot
Outcome example: Rotherham NHS DNA AI pilot Predictive engagement only matters when it changes the next best action. Own data ML model Extra SMS historic attendance signals risk stratification targeted 1 day before Pattern A ML prediction model trained on the organisation’s own data becomes a busine
Slide 8: Outcome example: Rotherham NHS ITSM Agent
Outcome example: Rotherham NHS ITSM Agent A repeatable agentic pattern: goal, tools, guardrails, outcome. 28% 41% GenAI NLU reduction in inbound IT calls handled by Autonomous Agent understands informal intent Pattern The agent has a goal, tools and defined guardrails. It resolves ITSM Tools Escalat
Slide 9: Outcome example: Manchester City Council
Outcome example: Manchester City Council Email automation: the repeatable IDP pattern hiding in plain sight. 1,004 20% 18% emails auto-triaged in week one tenancy emails auto-processed auto-replied: insufficient data Pattern Inbound email is read, understood, structured, indexed, routed and actioned
Slide 10: Outcome example: University Hospitals Sussex
Outcome example: University Hospitals Sussex Healthcare access: voice and digital engagement connected around the same service path. 86% <10% <3 mins digital engagement abandonment wait time Pattern Contact centre, patient portal, AI assistant, multilingual support, proactive 13% → 3–4% 1,500+ AI DN
Slide 11: Outcome example: Tewkesbury Borough Council
Outcome example: Tewkesbury Borough Council Local government: AI-enhanced citizen experience and measurable operational shift. 67,000+ 21,000+ 50% inbound calls processed emails processed wait-time reduction Pattern NLP and a local-government model interpret resident intent, route by service and 16.
Slide 12: A simple AI readiness scorecard
A simple AI readiness scorecard Use this before approving another pilot. Known demand A repeated contact reason or operational bottleneck. Baseline Current volume, cost, quality, and cycle time are visible. Operating context Customer, case, policy, status, and available actions. Bounded scope A cons
Slide 13: The antidote to AI hype
The antidote to AI hype AI that delivers results is built on operational discipline. No cool demos No silver bullet No magic dust The only demo that matters is AI will not fix a broken A bad process with AI is still a the measured service journey by itself. bad process but faster. outcome. The progr