Skip to content

Customer service is changing

Companies have always defined service on their terms.
Now customers are defining it on theirs.

Consumer AI is moving fast, and it's already in this space.

We wrote the Service Handshake — an open standard for how service works when one or more parties is an AI agent — and we build the services that work for both sides.

The Service Handshake — open standard

We wrote the open standard for services where one or more party is an AI agent.

Tested in live contact centres — human agents unaware they were talking to AI Published open standard — CC BY 4.0, citable, implementable today EU AI Act enforcement August 2026 — declarations map directly to compliance
Dual CX

Four modes. Most companies are designing for two.

Every customer service interaction now sits in one of these four modes. Modes 1 and 2 are designed. Modes 3 and 4 are already happening — and almost nobody is ready.

Human to human. High cost, high empathy, declining as a share of all interactions. Risk: over-investing while the world moves on.

A person contacts your company and interacts with your AI. Chatbots, voice AI, self-service. Designed for containment, not resolution. Every CCaaS vendor is selling the same thing.

A customer’s AI agent calls your contact centre. It passes verification. It explains the problem clearly. Your agent has no idea they’re talking to a machine. We have tested this. It works. Your operation isn’t designed for it.

Two AI systems interact directly. No human in the loop. Without declared goals, authority, and fallback rules, the interaction loops without resolution or causes active harm. The Standard defines how to design this safely.

The engine

One system. Built for all four modes.

Every interaction starts with a declaration. The engine operationalises it. The outcome is a service that works whether the customer is a person, an AI agent, or both.

Every interaction starts with a declaration

Before anything happens, each party declares their terms. These six elements come from the Service Handshake.

Goals Options & constraints Data permissions Fallback rules Cost parameters Authority level
What powers it
Knowledge packs AI models Processes & policies Load balancing Data & reporting Tools & integrations Workforce management
What the engine delivers
Structured knowledge AI agents built for resolution People who run both sides Aligned outcomes
The difference

Who’s ready for what’s coming?

Three approaches to service. Only one is designed for all four modes.

Traditional model AI-first model Neos Wave
Mode 1 Human ↔ Human Designed for this — but expensive to scale Deprioritised — seen as legacy Staffed with people who stay and compound
Mode 2 Human ↔ Brand AI Bolted-on chatbot, optimised for deflection Strong here — but generic AI, no domain depth AI fed by elevated knowledge — resolves, not deflects
Mode 3 Customer AI ↔ Human No awareness this is happening No awareness this is happening Designed for it — team trained, protocols in place
Mode 4 Customer AI ↔ Brand AI Not on the roadmap Not on the roadmap Defined the standard. Serving early cases.

Mode 1 — Human ↔ Human

TraditionalDesigned for this — but expensive to scale
AI-firstDeprioritised — seen as legacy
Neos WaveStaffed with people who stay and compound

Mode 2 — Human ↔ Brand AI

TraditionalBolted-on chatbot, optimised for deflection
AI-firstStrong here — but generic AI, no domain depth
Neos WaveAI fed by elevated knowledge — resolves, not deflects

Mode 3 — Customer AI ↔ Human

TraditionalNo awareness this is happening
AI-firstNo awareness this is happening
Neos WaveDesigned for it — team trained, protocols in place

Mode 4 — Customer AI ↔ Brand AI

TraditionalNot on the roadmap
AI-firstNot on the roadmap
Neos WaveDefined the standard. Serving early cases.
How we deliver

Three ways to run the engine

Our engine combines AI, people, knowledge, and process to deliver service in any of the four modes. Same system, same standard — you choose the relationship.

We add capacity

Support

Your platform, our people and engine. Mode 1 and 2 coverage with Mode 3 readiness built in. Deployed when you need it, standing down when you don’t.

Product launches · Enterprises · Regulated sectors
We work inside

Augment

Our people and AI work inside your stack. All four modes covered, shared operation. Better results, same tools, lower cost than dedicated headcount.

Switching BPOs · D2C · SaaS · Scale-ups
We own the outcome

Adaptive

We run it. Every mode, end to end. You pay on resolution, not headcount.

Start-ups · Scaling brands · New market entry

We only price on outcomes in Adaptive because that’s the only model where we control everything that drives one. If the knowledge isn’t right, we fix it. If the AI fails, our team catches it. You pay when the customer’s problem is solved.

Talk to us about Adaptive →
Support
Augment
Adaptive
For people

Companies have always had people to handle this.
Now so do you.

We customise AI agents to deal with companies so you don’t have to. The same engine we build for brands — now assembled for people.

Time back

Bills, renewals, cancellations, complaints — the life admin that eats your evenings. Your assistant handles the conversations. You get the outcome.

You stay in control

Your assistant acts within the limits you set. It can negotiate your broadband bill — but it can’t cancel your contract. You define the boundaries. We call this a Service Handshake.

Loved ones supported

Set up an assistant for a parent or family member. It handles the conversations with providers on their behalf — and keeps you in the loop. Triadic authority, designed properly.

The Service Handshake in action

AI that knows its limits

A family member sets up an AI assistant for their dad. The assistant can research and compare — but it can't commit to anything without approval. Watch what happens when Dad wants to switch broadband.

Dadbot
Powered by the Service Handshake

Want your own AI assistant?

What would you want help with? (pick any that apply)

No spam. We’ll let you know when it’s ready.

Who we are

Built by people who’ve done it. Writing what comes next.

Neos Wave exists because service operations weren’t designed for what’s coming. We’ve spent over two decades building and running CX at scale — and the past year testing what happens when AI agents enter the conversation. The Service Handshake standard came from that work. The engine came from everything before it.

25+ years building CX at scale
95% of work from recommendation
<5% team attrition — industry averages 30%
18 countries — global reach, local service

Brands that trust our team

Retail & Ecommerce ASOS, Selfridges, Kiko Milano, DFYNE, Karen Millen, Oasis, Cult Beauty, Laced, Dunelm, Monsoon
Tech & Finance Spotify, Atom Bank, Funnel, Openlane, Vodafone, ROLI, Melody VR
Lifestyle & Luxury Zegna, Bugaboo, Audi, Mercedes AMG Petronas, Years
Public, Social & Health Peabody Housing, Gravesham Borough Council, Pfizer

We wrote the standard. We build the implementation. We are not a consultancy, a BPO, or a software vendor. We are the team that defined what comes next.

Let’s talk.

Talk to AI Maria — she can talk about Neos Wave, the Service Handshake, Dual CX, and how we work. Or leave a message.

Talk to AI Maria

Ask about Neos Wave, the Service Handshake, Dual CX, or how we work.

Prefer to write? Leave us a message.