In Q2 2026, Forrester Research published its first landscape report on what it calls Adaptive Process Orchestration — a newly defined market category covering platforms that blend AI agents and nondeterministic control flows with traditional deterministic automation to execute complex business processes at scale. The report surveyed 35 vendors: Appian, ServiceNow, Camunda, UiPath, Workato, IBM, Automation Anywhere, Salesforce, Boomi, and 26 others operating across the category.
The number one barrier to adoption Forrester identified was not a technical limitation. It was not cost, integration complexity, or model reliability. It was this: enterprises have not done enough to reduce the trust barrier. Specifically, limited APO adoption stems from lack of AI trust and IP protection concerns.
That is a governance problem. And it is structural — baked into how APO platforms are architected, not addressable with a feature update.
What Adaptive Process Orchestration Actually Is
APO software platforms combine AI agents with both nondeterministic and deterministic control flows to meet business goals, perform complex tasks, and make autonomous decisions. The practical shift this represents is significant: organizations are moving away from brittle chains of RPA bots and rigid sequential workflows toward systems where AI agents can reason, adapt, and take actions that were not explicitly scripted in advance.
Forrester identifies four core use cases driving APO adoption: complex end-to-end process orchestration, agentic process orchestration, legacy modernization, and process execution in highly regulated environments. That last category — regulated environments — is where the governance gap is most acute and most consequential.
Extended use cases are also emerging at the frontier of the market. One that stands out: orchestration as an MCP service, where fully orchestrated processes are exposed to third-party AI assets through an MCP-compatible interface. This is no longer a theoretical architecture — it describes a real deployment pattern appearing in production agentic systems today.
There are, by Forrester's count, more than 400 copilot and agent building systems on the market. Most do not suit long-running, complex processes. The APO category is specifically for the workflows that are consequential enough — and long-running enough — that getting governance wrong has real organizational and regulatory consequences.
The Structural Problem: Governance Is Not a Feature of Orchestration
The 35 vendors in Forrester's landscape are, first and foremost, orchestration platforms. They handle process design, workflow execution, integration management, and increasingly, agent deployment. Some have added governance-adjacent features — audit logging, role-based access controls, rudimentary policy settings.
But governance is not a feature of orchestration. It is a separate architectural plane.
Here is why this matters: an orchestration platform that also enforces governance has a fundamental conflict of interest built into its architecture. The system responsible for running a process cannot simultaneously be the independent authority that determines whether that process should run, what constraints apply at each step, and whether outputs satisfy safety and compliance requirements. That is not a design choice a vendor can engineer around — it is a limitation of monolithic architecture.
Effective AI governance must sit outside the orchestration layer. It needs to intercept before execution, monitor during execution, and verify after execution — operating as an independent control plane that the orchestration system cannot override. When governance is a module inside an orchestration platform, it is always subordinate to the system it is supposed to constrain.
Forrester's research surfaces this tension directly. The functionality analysis for "process execution in highly regulated environments" lists governance hub, runtime monitoring and control, rules engine, roles and access management, and fail-safe operational features as primary requirements — not secondary considerations. These are load-bearing capabilities. They are the difference between a platform a regulated enterprise will deploy and one it will not touch.
The gap is that the APO platform market, as a category, does not natively close all five of those requirements at depth. Individual vendors cover subsets. None are dedicated governance control planes. That is the structural hole in the landscape.
The Orchestration Washing Problem Makes It Worse
Forrester identifies "orchestration washing" as the primary challenge in the APO market: many automation vendors swapped the word "automation" for "orchestration" as they added surface-level AI capabilities to existing products. The result is a market where 35 vendors claim APO capabilities, but buyers have no reliable mechanism to distinguish genuine orchestration and governance depth from repackaged point solutions with an AI label.
This is not a minor nuisance. It is the mechanism that stalls enterprise adoption. When a CTO cannot confidently evaluate whether a vendor's "governance hub" is a real pre-execution policy enforcement engine or a renamed audit log setting, the default answer is not to pick the right vendor — it is to delay the deployment. That is where the 400-plus agent building and copilot systems in the market leave enterprise buyers: overwhelmed, skeptical, and slow to commit.
The solution is not a better vendor evaluation rubric. The solution is an independent governance layer that operates across orchestration platforms — one that does not require the buyer to trust a vendor's self-reported governance claims, because governance is enforced externally by a dedicated control plane, regardless of which APO platform runs underneath.
How Waxell Addresses the APO Governance Gap
Waxell is not an APO platform. It does not compete with Appian, Camunda, ServiceNow, or the other vendors in Forrester's landscape. Waxell is the governance control plane that makes APO adoption viable — specifically in the regulated enterprise environments where the gap identified in that landscape is most consequential.
Three products, each addressing a distinct layer of the problem:
Waxell Observe instruments AI agents at the runtime layer, capturing every LLM call, tool invocation, input, output, and decision trace across 200+ libraries automatically — initialized with two lines of code. This is the visibility layer that makes "runtime monitoring and control" a real capability rather than a dashboard that surfaces what already happened. Observe instruments agents independently of which orchestration platform is running them, giving teams continuous signal across their entire agentic process estate.
Waxell Runtime enforces 26 policy categories at the pre-execution, mid-step, and post-completion stages of every agentic workflow — before an agent takes an action, not after. Policy categories include PII handling, scope constraints, cost hard stops, prompt injection detection, output validation, and human-in-the-loop escalation triggers. For process execution in regulated environments, this is the governance hub that Forrester's analysis treats as a primary differentiator. Runtime sits outside the orchestration platform, wrapping it, enforcing constraints the orchestration layer cannot self-impose. Policy enforcement details are documented at waxell.ai/capabilities/policies.
Waxell Connect governs the agents the team did not build: vendor agents, third-party integrations, and MCP-native agents operating within or alongside orchestration workflows. No SDK required, no code changes to the agent itself. As orchestration-as-an-MCP-service becomes an established deployment pattern — and Forrester's research confirms it is an emerging extended use case — Connect provides the policy enforcement and audit layer for every MCP call crossing organizational and vendor boundaries. Agent inventory and registry management are documented at waxell.ai/capabilities/registry.
Together, these three products address the architectural gap that Forrester's landscape surfaces but that the APO platform category does not natively close: independent, external, cross-platform governance for agentic process automation at enterprise scale. An overview of how the three products work together is available at waxell.ai/overview.
Where the APO Market Goes Next — and Why Governance Leads
Forrester's forward-looking analysis in the landscape report makes a prediction worth underscoring: once the current phase of buyer confusion clears, differentiation in the APO market will shift away from commoditized orchestration and agent-building capabilities toward specific context, governance, and deep industry expertise.
This is the direction every maturing software category converges on. Build-and-run capabilities get absorbed into platforms and eventually into infrastructure defaults. The durable differentiator becomes the governance layer — the system that tells teams what their agents are doing, enforces the constraints their industry requires, and produces the audit evidence compliance teams can stand behind in regulatory examinations.
Organizations in financial services, healthcare, insurance, and legal services are not waiting for the APO market to mature before facing regulatory expectations for AI governance. The EU AI Act (now in phased enforcement, with high-risk system obligations under Annex III taking effect August 2, 2026), SEC examination expectations for algorithmic AI systems, and HIPAA obligations for AI in clinical workflows are active compliance considerations today, not roadmap items for 2027. Enterprises in these verticals need a governance control plane now — independent of whichever orchestration platform they choose.
That is the gap. That is what Waxell is built to fill.
FAQ
What is adaptive process orchestration?
Adaptive process orchestration (APO) refers to automation platforms that combine AI agents and nondeterministic control flows with traditional deterministic workflow logic to execute complex, multi-step business processes autonomously. Unlike legacy robotic process automation, which follows rigid scripted sequences, APO systems can reason, adapt to changing conditions, and pursue business goals without requiring every step to be explicitly defined in advance. Forrester Research formally defined and named this market category in its Q2 2026 landscape report covering 35 vendors.
Is Waxell an APO platform?
No. Waxell is a governance control plane for agentic systems, not a process orchestration platform. Where APO platforms handle process design, workflow execution, and agent deployment, Waxell provides the independent governance layer that sits above and across those platforms — enforcing policies before execution, monitoring runtime behavior continuously, and governing third-party and vendor agents regardless of which orchestration system is running them. Waxell does not replace APO platforms; it makes their enterprise deployment viable.
What is the governance gap in adaptive process orchestration?
The governance gap refers to the structural absence of independent, external governance in the APO platform category. Many APO vendors include governance-adjacent features — audit logs, access controls, policy settings — but these are internal to the orchestration system itself. Effective governance for regulated environments requires a control plane that sits outside the orchestration layer, enforcing constraints before and during execution rather than logging what happened afterward. Forrester's Q2 2026 APO landscape identifies this gap implicitly: it lists governance hub and runtime monitoring and control as primary requirements for regulated-environment use cases — capabilities the APO market as a whole does not provide at dedicated-layer depth.
What is orchestration washing?
Orchestration washing describes the practice of automation vendors relabeling existing products as orchestration or APO platforms after adding minimal AI capabilities. The term was surfaced in Forrester's Q2 2026 APO landscape as the market's primary challenge: buyers cannot reliably distinguish platforms with genuine orchestration and governance depth from repackaged point solutions, which slows enterprise adoption across the category. The practical consequence is that enterprise buyers delay deployment rather than risk selecting a platform whose governance claims they cannot verify independently.
How does governance enable APO adoption in regulated industries?
Regulated industries — financial services, healthcare, insurance, legal — require audit trails, policy enforcement, data handling controls, and compliance evidence before they will deploy autonomous AI systems at scale. Without a dedicated governance layer, APO platforms cannot provide the independent verification that regulated enterprises require. Governance addresses the trust deficit Forrester identifies as the primary barrier to APO adoption: once enterprises can demonstrate that agentic workflows operate within defined constraints and produce auditable records, deployment velocity increases. The governance layer is not a compliance checkbox — it is the architectural prerequisite for production deployment in regulated environments.
What is the difference between runtime monitoring and governance?
Runtime monitoring is visibility — it shows what agents are doing as they do it. Governance is enforcement — it determines what agents are permitted to do before they act, and stops or escalates when constraints are violated. Monitoring is necessary but not sufficient for compliance. A dashboard that logs an agent's unauthorized data access after the fact is not governance; it is forensics. Waxell Runtime enforces policies across 26 categories at the pre-execution stage of every agentic workflow. Waxell Observe provides the continuous runtime monitoring layer that feeds signals into that enforcement. Both are required; neither substitutes for the other.
Sources
- Forrester Research. The Adaptive Process Orchestration Software Landscape, Q2 2026. Bernhard Schaffrik and four contributors.