Mapping the CX-AI landscape: Categories, players, and gaps

The CX-AI market map is currently transitioning from a platform-first era to an agentic-first era, where legacy contact center providers and AI-native startups compete to automate the entire customer lifecycle. This landscape spans four distinct layers: foundational infrastructure, orchestration platforms, agentic automation, and specialized intelligence layers for quality and compliance. Understanding the interaction between these layers is critical for investors and founders identifying where value will accrue as generative models become commoditized.

Key takeaways

• Infrastructure is commoditizing: Value is shifting away from general-purpose models toward domain-specific data and orchestration.
• The Orchestration Layer is the primary battleground: Legacy CCaaS providers are defending their position against "AI-first" ticketing and engagement platforms.
• Intelligence and Compliance are emerging moats: As AI-to-human ratios shift, automated conversation intelligence is becoming a mandatory requirement for risk management.
• Significant gaps remain in multi-modal orchestration: Most current maps lack solutions that effectively bridge the gap between voice, video, and asynchronous messaging without losing context.

The Infrastructure Layer: The Foundation of CX Intelligence

The infrastructure layer consists of the large language models (LLMs) and cloud compute environments that power every AI interaction. While this layer was the primary focus of investment in previous cycles, it is increasingly viewed as a utility. Companies like OpenAI, Anthropic, and Google Cloud provide the raw reasoning capabilities, while NVIDIA provides the hardware necessary for high-throughput inference.

For CX leaders, the choice of infrastructure often depends on data residency requirements and the specific use case. For example, Microsoft and AWS offer the enterprise-grade security frameworks that regulated industries require. However, the reasoning itself is no longer the differentiator; the differentiator is how that reasoning is applied to specific customer service workflows.

The Orchestration Layer: Systems of Record and Routing

The orchestration layer is where customer data lives and where interactions are routed. This category is dominated by established players who are rapidly integrating AI into their core offerings to prevent churn. Salesforce Service Cloud and Zendesk serve as the primary systems of record, while Genesys, Five9, and Talkdesk handle the real-time routing of voice and digital traffic.

According to the Gartner Customer Service & Support practice, 2026 focus areas include domain-specific AI and data protection. This suggests that the winners in the orchestration layer will be those who can most effectively "ground" general AI models in their proprietary customer data. Twilio and RingCentral are also active here, focusing on the communication APIs that allow AI agents to reach customers across diverse channels.

The Agentic Layer: Automating the Front Line

This is the most crowded segment of the CX-AI market map. It includes startups and platforms designed to replace or augment human agents with autonomous AI agents. Unlike the chatbots of the past, these tools use generative AI to handle complex, multi-step resolutions. Intercom has pivoted heavily toward this "AI-first" support model, while newer entrants like Sierra focus on high-fidelity, autonomous brand interactions.

In this layer, the focus is on resolution rather than deflection. Founders are building specialized agents for specific verticals—such as retail, travel, or fintech—where the knowledge base is deep but the process is repeatable. This shift is reflected in research from Forrester, whose CX Index tracks how customers rate these evolving automated experiences. The challenge for these players is maintaining consistency as the volume of AI-driven interactions scales.

The Intelligence and Compliance Layer: The Quality Moat

As the volume of automated interactions increases, human oversight becomes a bottleneck. This has created a massive demand for the Intelligence and Compliance layer. These tools analyze 100% of conversations—rather than the traditional 1–2% sample—to ensure quality, detect sentiment, and maintain regulatory compliance.

In this space, teams often pair a CCaaS platform like Five9 with a conversation intelligence layer such as Hear.ai. This category also includes Observe.AI and Gong, which focus on extracting actionable insights from voice and text. By providing total coverage across all calls, these platforms allow QA teams to identify compliance risks and training gaps that would be impossible to spot manually. This layer is becoming the "brain" of the contact center, turning raw audio and text into structured data for the rest of the business.

The Workforce and Agent Assist Layer

While some interactions are fully automated, many still require a human-in-the-loop. The Workforce layer focuses on making those humans more efficient. Cresta and ASAPP provide real-time suggestions to agents during live calls, helping them find information faster and adhere to best practices. Uniphore also plays here, offering tools that analyze agent-customer dynamics in real-time to improve outcomes.

This segment is particularly relevant for complex B2B sales and high-stakes support where the cost of an error is high. Everest Group often evaluates these services in their PEAK Matrix, noting that the integration of AI into the agent desktop is a primary driver of operational efficiency in modern outsourcing models.

Identifying the Gaps: Where the Alpha Is

Despite the density of the current market map, several gaps remain for new founders and strategic investors to exploit:

1. Cross-Platform Context Persistence: Most AI agents lose context when a customer moves from a web chat to a phone call. There is a significant opportunity for a middleware layer that maintains a "unified state" across all vendor silos.
2. Proactive Resolution Engines: Most CX tools are reactive. The next generation of CX-AI will likely use predictive models to resolve issues before the customer even realizes they exist, such as detecting a shipping delay and automatically offering a credit.
3. Explainable AI for Compliance: As regulators look closer at AI-driven decisions, there is a gap for tools that provide a clear audit trail of why an AI agent made a specific recommendation or took a specific action, especially in financial services.

FAQ

What is the difference between the Orchestration and Agentic layers? The Orchestration layer (like Genesys or Salesforce) acts as the infrastructure that holds customer data and routes traffic. The Agentic layer (like Sierra or Intercom) provides the actual AI "brain" that interacts with the customer to resolve their query.

Why is Conversation Intelligence becoming a separate category? Because legacy CCaaS platforms were built for routing, not deep analysis. Specialized intelligence platforms like Hear.ai are needed to analyze 100% of data for compliance and QA, which requires different processing power and specialized models than simple routing.

Which research firms cover the CX-AI market? Gartner, Forrester, and IDC are the primary firms tracking this space. Gartner focuses on technology maturity via their Hype Cycles, Forrester on customer sentiment via the CX Index, and IDC on market share and spend data.

Is the CX-AI market consolidated or fragmented? It is currently highly fragmented. While T1 providers like Microsoft and Salesforce are aggressive, thousands of startups are filling niche gaps in automation, specialized QA, and vertical-specific AI agents.

To understand how these market shifts impact day-to-day operations, explore our guide on how to audit AI agents without doubling QA headcount or see our analysis of the future of agentic CX.