Extole Evolves Its Developer Platform for AI-Assisted Teams
Referral and loyalty platforms were built for users. The next generation of software will also need to work for developers, automation systems, and AI agents.
As AI becomes part of everyday operations, teams increasingly expect their platforms to be programmable—not just configurable. Developers want APIs they can build against. Operations teams want workflows they can automate. AI assistants need structured interfaces they can understand and safely operate.
That’s why we’ve significantly expanded Extole’s developer platform.
With a redesigned OpenAPI-based API surface, an expanded command-line interface, and new AI-ready access through the Extole MCP, Extole is evolving from a platform teams use into infrastructure they can build on. Developers can integrate more deeply, automate more confidently, and give AI tools access to the same platform capabilities through secure, governed interfaces.
These updates meaningfully expand what you can do with Extole, and make it easier and faster to actually build it.
Three Surfaces, One API Layer
Whether you’re working in code, in a terminal, or through an AI assistant, you’re interacting with the same underlying Extole platform.
The top priority for our product team was exposing more of the platform architecture to devs and AI agents, giving them a complete picture of how each component operates. The next important step was providing AI-enabled tools to let developers build, test, launch, manage, and optimize offer programs through Extole.
The expansion is best understood as three connected surfaces sharing the same underlying API:
MCP
The Extole MCP connects Extole to any MCP-compatible AI environment — Claude, Cursor, ChatGPT, and others. Developers, marketers, and AI agents can use plain language to inspect programs, retrieve data, update configurations, and operate offer programs without logging into My Extole.
CLI
The Extole CLI brings the same surface to the terminal. Script it, pipe it, integrate it into CI, or use it as the interface through which agents take direct action on the platform. Because the CLI and MCP share the same underlying API surface, workflows built in one are portable to the other.
Expanded API
Expanded OpenAPI-compliant APIs are what make both work. Extole’s API surface has been significantly restructured: full program control through a single coherent interface, strict adherence to OpenAPI standards so your IDE and AI tools know exactly what is available, and substantially improved reference documentation. If you’ve connected to Extole’s APIs before and found the surface inconsistent or hard to navigate, the current state is materially better.
The design principle behind all three: if a developer can understand it clearly, an AI agent can use it reliably.
Building on Extole as Infrastructure
Together, the API, CLI, SDKs, and MCP create a consistent developer surface across the entire platform. Whether teams are integrating customer events, configuring rewards, managing campaigns, or analyzing performance, the same underlying model is available through every interface.
Component-Based Architecture
Extole’s component-based architecture — events, audiences, rules, offers, rewards, messages, journeys, integrations, and reporting — is what makes these surfaces useful rather than decorative. Because the platform is configurable and not template-bound, the API, CLI, and MCP expose genuine program control, not a thin wrapper around a fixed feature set.
Highly Customizable Programs
A developer building on Extole can configure custom eligibility flows, define audience-specific reward logic, integrate backend event streams, and connect to external CRMs, CDPs, and commerce platforms — all through the same surfaces the MCP and CLI access. The AI-readable API layer means agents can traverse component relationships and understand the platform’s structure from the JSON itself, without requiring platform-specific training.
Faster and More Secure Than DIY
For teams evaluating whether to build incentive infrastructure themselves or buy a platform, this is the relevant comparison point. Extole provides the offer, reward, fraud, fulfillment, and audit infrastructure that would otherwise require months of engineering to build and maintain. The expanded developer surface makes the platform faster to evaluate, faster to integrate, and faster to extend.
What the API Expansion Actually Covers
Extole’s three API layers — Server to Extole, Consumer to Extole, and the Management API — have been documented and structured consistently under OpenAPI contracts. The result is less time spent discovering platform behavior and more time building customer-facing experiences.
The Server to Extole API handles event submission from your backend and runs bulk data operations.
The Consumer to Extole API powers browser and native app experiences via JavaScript, iOS, Android, and React Native SDKs.
The Management API lets you configure programs in code the same way you would in My Extole — managing campaigns, audiences, reward suppliers, and reporting through a single, stable interface.
Every published endpoint ships an OpenAPI definition. Your IDE can generate clients from it. Your agents can read it directly. For teams that previously had to reverse-engineer behavior or rely on internal documentation to understand how components related, this is a genuine step change in how quickly you can evaluate, integrate, and extend the platform.
The API documentation is published on ReadMe and GitHub, and the full OpenAPI spec is available for use in Postman and agent tooling.
Operating Extole Through AI Tools You Already Use
The new Extole MCP is a remote control surface for the platform, accessible through whatever AI tool you already have open.
A developer using Cursor or Claude can inspect Extole’s full component structure and ask the agent to explain how components relate before writing a single line of integration code. Configuration work that previously required back-and-forth between engineering and solutions teams can move substantially faster. One internal example: a complex SSO integration that would typically take multiple review cycles was completed in approximately 30 minutes with MCP-assisted configuration.
For marketers, the MCP covers the operational queries that take up time for marketing, engineering, and support teams: what changed, what is active, who qualified, which rewards were issued, and how a campaign is performing. Marketers can also now manage, track, and edit programs through the MCP using plain language, simplifying the time it takes to update reward values, pull reports, and more.
The MCP uses OAuth 2.0 with scoped tokens, so agents operate with exactly the permissions you grant them. Every action is logged in Extole’s change log with the originating user, tool name, and timestamp, giving program governance teams a full audit trail.
AI-Assisted Configuration, Deterministic Execution
There is an architectural boundary in how Extole’s MCP works that matters to any team running incentive programs, where fraud and gaming are always a risk.
AI belongs in the configuration and validation layer. That is where it adds the most value: helping teams draft program structures, inspect component relationships, identify configuration gaps, and move faster from intent to launch. These are tasks where iteration is fast, and correction is easy.
What AI should not own is runtime execution. With offer programs, that refers to eligibility decisions, fraud detection, reward authorization, and fulfillment. These are not tasks where probabilistic reasoning is an acceptable tradeoff — with $3.1 billion in fraudulent loyalty redemptions annually, the system handling fraud and reward delivery needs to be deterministic.
Extole was designed so that AI can help operate programs without becoming responsible for executing them. When an agent operates through the MCP, it works at the configuration layer. The runtime underneath remains Extole’s own deterministic infrastructure: sub-200ms event processing, real-time eligibility and quality rules, live fraud detection, reward authorization, and fulfillment.
The MCP permission model reinforces this, providing additional guardrails to prevent unauthorized program access.
What Teams Can Do With This Today
Our revamped developer center and AI tools are for more than just engineers—they are designed to enable easier, faster operations across the entire program lifecycle. Here is what each one of your teams can do now that the MCP, CLI, and API expansion are in place.
Developers can:
- Validate a campaign configuration before launch
- Pull reward issuance history across programs
- Compare audience targeting changes over time
- Automate configuration checks in CI
Marketers and program managers can:
- Generate a report from campaign performance data
- Build a welcome offer flow
- Review participant and reward history
- Change targeting rules for a campaign
- Edit a reward value for a campaign
Operations and analytics teams can:
- Pull conversion data for a specific campaign
- Summarize campaign performance across accounts
- Review campaign change logs to trace configuration changes
- Check program status and monitor overall program health
All of this is possible through plain-language requests in Claude, ChatGPT, or any MCP-compatible client. Internal beta usage has shown that the tool is already highly skilled at understanding and explaining campaign structures, including summarizing integrations, reviewing reward distribution patterns across segments, and analyzing changes in journey targeting over time.
Getting Started
The Extole MCP is available now for customers across developer, marketer, and operations workflows. The CLI and expanded OpenAPI-compliant APIs are part of the same release, accessible through the Extole developer center and documented on ReadMe and GitHub. Extole’s AI roadmap continues from here — pre-publish validation for components, expanded write operations, and deeper integration with agent frameworks are on the near-term roadmap.
If your team is building referral, reward, loyalty, or lifecycle incentive programs and you want infrastructure your developers can integrate confidently and your AI agents can operate safely, this is the place to start. Request a demo to see what building on Extole looks like for your stack.
Frequently Asked Questions
What is a developer platform?
A developer platform provides APIs, tools, and infrastructure that developers can build on and integrate with their applications. Extole’s developer platform includes OpenAPI-compliant APIs, a command-line interface, and an MCP that lets developers, operations teams, and AI agents configure and manage referral and loyalty programs programmatically.
What is the difference between a developer platform and developer portal?
A developer platform is the underlying infrastructure and APIs that power software development, while a developer portal is the interface where developers access documentation and interact with those tools. Extole provides both: the platform infrastructure for referral and loyalty programs, plus the Extole MCP and CLI as accessible interfaces to operate that infrastructure.
How does Extole’s MCP work with AI tools?
The Extole MCP connects to any MCP-compatible AI environment, such as Claude, Cursor, or ChatGPT, letting users inspect programs, retrieve data, and update configurations using plain language. It operates at the configuration layer with OAuth 2.0 scoped permissions, while Extole’s deterministic runtime handles actual reward authorization and fraud detection.
What can teams do with Extole’s expanded API surface?
Developers can validate campaign configurations, pull reward history, and automate checks in CI; marketers can generate reports, build offer flows, and edit reward values; operations teams can analyze campaign performance, review change logs, and monitor program health. All three surfaces—API, CLI, and MCP—access the same underlying platform capabilities.
What is included in Extole’s developer platform expansion?
The expansion includes three major components: a redesigned OpenAPI-compliant API surface, an expanded command-line interface (CLI), and AI-ready access through the Extole MCP. Together, these capabilities make it easier to integrate, automate, and manage referral, loyalty, and incentive programs.
What APIs does Extole provide?
Extole offers three primary API layers: the Server to Extole API for backend event submission and bulk operations, the Consumer to Extole API for web and mobile experiences, and the Management API for configuring campaigns, audiences, rewards, and reporting.
Can developers automate Extole workflows?
Yes. Developers can use the Extole CLI and APIs to automate configuration checks, integrate Extole into CI/CD pipelines, retrieve reporting data, validate campaign configurations, and build custom workflows around incentive programs.
How does Extole support AI-assisted program management?
Through the Extole MCP, AI assistants can help teams inspect program structures, review configurations, generate reports, analyze performance, and identify potential issues. This reduces manual effort while maintaining governance and control.
Is Extole secure for AI-assisted workflows?
Yes. The Extole MCP uses OAuth 2.0 with scoped permissions, ensuring AI tools only have access to approved resources. All actions are logged with user, tool, and timestamp information to provide a complete audit trail.
Can AI agents execute rewards or fraud decisions in Extole?
No. Extole uses AI for configuration and validation workflows, but runtime functions such as eligibility decisions, fraud detection, reward authorization, and fulfillment remain deterministic and are handled by Extole’s core infrastructure.
