One Engine, Six Deployments
There's a housing association in Kent that sends proactive WhatsApp messages to tenants about repairs, rent, and anti-social behaviour. The messages are generated by AI, reviewed by a human team, and dispatched at the right moment based on case data. Tenants reply. The AI responds. When the conversation needs a person, a person steps in — already briefed, already in context.
There's an automotive client whose back-office operations — warranty claims, dealer queries, parts logistics — are handled by a team that sits inside an AI-assisted workflow. Every query is triaged, enriched with knowledge, and routed before a human touches it.
There's a clinic that needed out-of-hours cover. Patients call after 6pm and reach an AI voice agent that handles appointment queries, symptom triage routing, and prescription repeat requests. The agent knows the clinic's protocols because it was trained on them.
There's a voice coaching product — live sessions between a human coach and a client, with AI providing real-time prompts, session notes, and follow-up actions.
There's a company that outsources its entire customer operation to us. We run the people, the AI, the knowledge base, the reporting, and the strategic layer. They get outcomes. We handle everything underneath.
And there's a consumer product — a personal assistant for someone's dad, built to handle broadband switching, bill comparisons, and service complaints on his behalf.
Six deployments. One engine.
Configuration, not architecture
The thesis behind Neos OS is simple: every service deployment is an assembly of the same components. Knowledge. AI. People. Reporting. The difference between a housing tenancy service and a personal assistant for a retiree is not the architecture — it's the configuration.
Which knowledge base? Housing regulations and tenancy policies, or broadband tariffs and switching procedures. Which tone? Formal and precise, or warm and patient. Which system of record? A CRM, a property management platform, or a WhatsApp thread. Which escalation path? A housing officer, a clinic receptionist, or a daughter who needs to approve the broadband switch.
The engine doesn't change. The parameters change.
This matters because most service companies are stuck in one of two moulds.
Platforms versus people
The first mould is the technology platform. SaaS companies that sell tools — ticketing systems, chatbot builders, analytics dashboards. They give you the raw materials but not the assembled service. You still need to hire agents, build knowledge, train the AI, and run the operation. The platform is necessary but not sufficient.
The second mould is the BPO. Business process outsourcers that sell people — headcount on seats, billed by the hour or the interaction. They provide labour, but the intelligence stays thin. Every new agent starts from scratch. Knowledge lives in the heads of the longest-tenured staff, and those staff leave at 30% per year. The BPO scales, but it doesn't compound.
Neos Wave is neither.
We don't sell a platform for you to operate. And we don't sell bodies to plug into your platform. We assemble AI, people, and knowledge dynamically for each deployment. The engine is ours. The configuration is yours.
Compounding
Here's what changes when you build it this way: every deployment makes the next one better.
The knowledge architecture we built for housing tenancy services taught us how to structure domain-specific AI training for regulated environments. That made the clinic deployment faster to ship. The voice agent we built for out-of-hours clinical cover taught us how to handle real-time protocol adherence. That fed back into the coaching product.
The consumer assistant — a personal AI for one person's dad — forced us to think about what service looks like when the 'customer' is a family and the 'agent' is an AI acting on someone's behalf. That thinking now shapes how we design every deployment, because Mode 3 and Mode 4 interactions are coming for all of them.
This is compounding. Not the startup cliché. The operational reality: every new deployment adds knowledge, patterns, and capabilities that the engine absorbs and the next deployment inherits.
What this is not
It's not a chatbot company. Chatbots are a feature, not a business. We build whole operations.
It's not a call centre. Call centres sell time. We sell outcomes.
It's not a consultancy. Consultants describe the future and leave. We build it and run it.
It's not even a tech company, really. The technology is essential but it's not the product. The product is a service that works — for a housing tenant chasing a repair, for a patient calling after hours, for a dad who wants cheaper broadband, for a brand that needs its warranty claims handled properly.
One engine
We started with the belief that the same engine could serve both sides of the conversation — the company and the customer. Six deployments later, that belief is operational fact.
The question we keep asking: what's deployment seven?