By Geoff McDonald, CEO & Co-Founder, Ambassador May 2026
Two years into the AI agent boom, the limitation of every customer-facing agent has become obvious.
The agent does not know who its customer is.
It knows what was typed into the chat window. It does not know what that customer bought last quarter, what campaign they responded to, whether they referred someone, what their lifetime value looks like, or what they are likely to do next. The customer relationship lives in one set of systems. The agent lives in another. Every interaction is essentially a cold start.
This is the gap we are building Ambassador to close.
Over the last 18 months we rebuilt the platform around a single thesis: the future of customer engagement is not more applications. It is a runtime that sits underneath every customer-facing agent and gives that agent the full context of who it is talking to, what has happened in the relationship, and what the right next move is.
We call this the Lifecycle Operating System for Customers and their Agents.
A Lifecycle Operating System is not a tool. It is the layer that makes every other tool smarter.
It has four parts:
Signal. Every customer touchpoint flows in. Subscriber data, network signals, billing events, support interactions, advocacy moments, lifecycle stage. One unified record per customer. We call this the Customer Outcome Graph.
Intelligence. A probabilistic scoring layer reads everything and ranks every customer for every outcome. Hiro, our orchestrator, predicts what is likely to happen next and what action is most likely to change it.
Action. When an agent fires (a chatbot, a campaign agent, a retention agent, a recovery agent), it does not start cold. It gets a script informed by the entire customer record. The right offer goes to the right person at the right moment.
Outcome. Every interaction is measured against the prediction. The system learns. The next prediction is more accurate. Compounding intelligence over time.
This is not a chatbot, a survey tool, or a referral platform. Those are applications. This is the runtime underneath all of them.
Three things made this the right direction.
Carriers, financial services, and B2B SaaS — the verticals where every customer interaction has revenue weight — are the verticals where AI agent deployments are accelerating fastest. These customers do not need another point tool. They need a runtime that connects everything they are already doing. We are already in production across all three: Verizon, Astound, Visible in carriers. H&R Block, CIBC, Oportun in financial services. Rippling, Dialpad in B2B SaaS.
Every well-funded AI company in 2026 falls into one of two camps.
The first is internal AI tools: Cursor, Glean, Copilot. These make employees more productive. They do not touch the customer.
The second is vertical AI applications: Sierra for customer service, Harvey for legal, Decagon for support. These solve one workflow. They sit at the application layer, not the runtime.
None of them are building the customer-facing runtime underneath. The lifecycle runtime layer for customer agents is open ground.
Companies are paying annual subscriptions to Medallia, SurveyMonkey, Qualtrics, Extole, Friendbuy, and a dozen others for tools they use ten percent of the time. The contracts compound. The data fragments. The customer record sits in seven systems and nowhere at once.
This is unsustainable, and the buyers know it. Every CFO conversation we have now starts with the same question: how do I stop paying for shelf-ware?
The answer is a consumption-based runtime. You pay for what you use. The platform fee covers the engines, agents, and orchestrator. Usage fees scale with channel volume. Consumption credits scale with AI execution. Three lines that compound together, with 105% net revenue retention on the new platform validating the model.
If you run customer programs, you are about to be asked to deploy AI agents into them. Some of you already are.
The question is not whether you will use agents. It is what those agents will know about your customers when they act.
If your agents run on an application stack, they will know what the user typed. If your agents run on a runtime stack, they will know everything you know about that customer. Across every touchpoint. Across every program. Across every channel.
This is the difference between an agent that completes a transaction and an agent that grows a relationship.
For our existing customers, this is why you are seeing the pace of releases. RBAC, default role sets, Agent Studio with Widget Creation and Campaign Intelligence already live, Automation Setup and Goal Setting rolling out this month, programmatic ads embedded mid-June, Customer Voice and Reviews behind that. Every release is a piece of the runtime. Each piece writes to the same Customer Outcome Graph. Each piece reads from the same record. The whole platform compounds.
For prospects looking at us for the first time, this is what we are. Not a referral platform that added some AI features. Not a survey tool with a chatbot bolted on. A runtime built from the ground up to give every customer agent the context it needs.
We are pointed at one thing.
Every customer touchpoint, on one record. Every agent that touches that customer, reading and writing to that record. Every outcome, predicted before it happens and measured against the prediction. Every dollar of customer activity, accounted for in real time.
That is the Lifecycle Operating System. That is what we are building. That is what our customers are using in production today, with 105% net revenue retention on the new platform and nine out of ten of our 2.0 customers actively migrating to 3.0.
If you are running a customer program in 2026 and you are not building on a runtime, you will be in 2027.
We are happy to be the runtime. Even happier when you bring your own agents and let ours feed them what they need.
That is what the next chapter of Ambassador looks like.
Geoff McDonald is CEO and Co-Founder of Ambassador. Ambassador is the Lifecycle Operating System for Customers and their Agents, in production across Verizon, Astound, H&R Block, CIBC, Visible, Rippling, and 200+ other brands. Learn more at getambassador.com.