Decagon Launches 'Proactive AI Agents' – Shifting Customer Support from Reactive to Preemptive

Decagon Officially Launches Its "Proactive AI Concierge"

Decagon Unveils Proactive Agents – The AI Concierge for Every Customer In The Agentic Commerce Era

Today Decagon, the leader in conversational AI agents for concierge customer experiences, introduced a new and proactive generation of its agentic technology...

On March 9, 2026, AI customer support company Decagon unveiled its next-generation "Proactive Agents." Unlike conventional passive AI chatbots that wait for customer inquiries, these agents anticipate customer needs and proactively reach out at the right moment — functioning as an "offensive AI concierge."

The two pillars of this new capability are "outbound voice" and "user memory," developed in collaboration with customers in the travel, retail, and health tech sectors.

Industry Trends

Despite being less than three years old, Decagon is a rapidly growing AI CX (customer experience) platform company. In March 2026, the company raised $250 million in Series D funding, reaching a $4.5 billion valuation. The round was led by Coatue Management and Index Ventures, with continued participation from a16z, Accel, Bain Capital Ventures, Forerunner, and Ribbit Capital.

Its customer base includes over 100 companies ranging from Fortune 100 enterprises to cutting-edge tech firms, such as Avis Budget Group, Block, Deutsche Telekom, Oura Health, Affirm, and Chime. The tripling of its valuation in just six months reflects the rapid expansion of agentic commerce — the emerging trend where AI agents act on behalf of consumers in purchasing decisions.

Across the e-commerce industry, major players like Mastercard, Visa, Google, and Salesforce have been rolling out agentic commerce initiatives. Against this backdrop, Decagon has moved to fully deploy proactive AI agents from the front lines of customer experience — the customer support domain.

Outbound Voice – The Era of AI Calling Customers

The first pillar of Proactive Agents is the "outbound voice" capability. Offered as part of Decagon Voice 2.0, this new feature enables AI agents to make outbound calls to customers on behalf of businesses.

Specific use cases include appointment reminders, onboarding progress check-ins, subscription renewal notifications, and loyalty reward suggestions. The goal is to transform contact centers from "inbound response hubs" into "proactive engagement engines."

On the technical side, Decagon has developed specialized voice models to address challenges unique to outbound calls, such as connection quality and answer detection. Voice 2.0 overall achieves a 65% improvement in generation latency, enabling streaming voice generation with natural prosody. Additionally, businesses can customize tone, pacing, and pronunciation to match their brand personality, configure response scenarios for emotionally sensitive situations, and set their own escalation rules for handoff to human agents.

Vikram Rajagopalan, VP of Customer Experience at rental car giant Hertz, commented: "Decagon has enabled us to shift from reactive support to proactive outreach, allowing our team to focus on interactions that require a human touch."

User Memory – How AI "Remembers" Customers

The second pillar is the "user memory" feature. This system accumulates conversation context, customer preferences, emotional signals, and behavioral patterns, then applies them to actual customer interactions.

The practical benefits are clear. For example, when a customer follows up on a troubleshooting issue from a few days ago, there's no need to explain everything from scratch. Because the AI agent remembers previously stated size preferences and feature requests, it can deliver personalized responses. Outbound outreach from businesses also becomes precisely targeted based on past engagement history, rather than blanket segment-based messaging.

Furthermore, the system features "cross-channel memory" that carries context across chat, voice, and SMS channels. If a customer starts an inquiry via chat and continues by phone, the conversation flows seamlessly within the same context. The system also complies with security and governance standards, balancing personalization with compliance.

At fintech company Chime, the system has achieved processing over 1 million calls per month with a 70% resolution rate and improved customer satisfaction.

Impact and Applications for E-Commerce Businesses

Decagon's move carries three important implications for e-commerce businesses.

Customer support is being redefined. Traditional CX centered on "handling inquiries," but we're entering an era where AI agents proactively reach out to customers. E-commerce-specific use cases such as proactive shipping delay notifications, cart abandonment follow-ups, and repeat purchase suggestions are expected to emerge.

The resurgence of voice channels deserves attention. With natural voice dialogue being added to what was previously text-chat-only AI customer support, reach is expanding — particularly to senior demographics and mobile users. The 65% latency improvement has brought quality to practical levels.

The granularity of customer data utilization is changing. With the user memory feature, not only structured data accumulated in CRM systems but also unstructured data such as conversation context and emotional cues can be leveraged in customer interactions. This has the potential to fundamentally elevate the precision of personalization.

Proactive Agents, including Voice 2.0, are now generally available (GA). For businesses considering adoption, the first step is to identify "touchpoints where proactive outreach should occur" within existing customer support flows.

Summary

Decagon's Proactive Agents represent a paradigm shift in AI customer support — from reactive responses to preemptive customer experiences. As the company, now valued at $4.5 billion with over 300 employees, scales its technology proven across travel, retail, and health tech, it has the potential to rewrite the standard for CX infrastructure in the agentic commerce era.

The key question going forward is how consumers will receive the experience of "AI agents calling them." Whether proactive outreach is perceived as "helpful" or "intrusive" hinges on the precision of timing and context. How far Decagon's user memory feature can elevate that precision, and the accumulation of real-world deployment cases, are worth watching closely.