Five Platforms, 72 Hours: AI Agents Now Table Stakes

Something happened this week that would have seemed unlikely even six months ago. Between June 1 and June 3, five separate companies — Meta, Cisco, Vonage, Klaviyo, and MoEngage — all shipped production AI agents for customer support, sales, and marketing. Not betas. Not waitlists. Actual products, available to businesses right now.

This is not a coincidence. AI agents for customer-facing work have moved from "emerging" to "expected."

Meta Business Agent: 1 million businesses, three messaging apps

Meta launched its Business Agent globally on June 3 at its Conversations conference in London. The agent handles customer responses, product recommendations, appointment booking, and lead qualification across WhatsApp, Messenger, and Instagram. Over one million businesses were already using it during pilot programs in India, Mexico, and Brazil. A billion people message businesses on these platforms every day.

The agent connects to Shopify, Zendesk, and Shopee through a new "Meta Business Agent Platform" that lets businesses customize and deploy at scale. It also delivers a morning briefing summarizing overnight chats. Getting started is free, with paid subscription tiers coming through Meta One.

This is probably the single largest AI agent deployment by raw business count. WhatsApp alone has 200 million small business users. Meta's paid messaging revenue hit a $2 billion annual run rate as of December, and the Business Agent is clearly part of pushing that number higher.

Cisco declares the chatbot era over

At Cisco Live in Las Vegas on June 2, the company unveiled three products under its Webex CX umbrella: AI WEM (Workforce Engagement Management), AI Concierge, and AI Agent 360.

AI WEM is a ground-up rebuild of workforce tooling for environments where human and AI agents coexist. Forecasting, scheduling, quality management, and performance dashboards all work across both agent types. AI Quality Management auto-evaluates 100% of interactions, not the sampled subset that legacy tools relied on.

AI Concierge is a pre-built front-door agent with persistent memory across channels. Vinod Muthukrishnan, VP and GM of Webex CX, described the goal as "no matter when I messaged you or called you, what the context was, what the modality was... you pick up and feel like we never stopped talking."

AI Agent 360 wraps observability, security, and management into a single control plane, drawing on Cisco's AI Defense and Splunk. Muthukrishnan was blunt: "Either do all of these things around security, observability, and what have you, or don't. There are no half measures."

That quote is worth sitting with. Cisco is betting that enterprises will pick a contact center platform based on whether it can secure and govern AI agents, not just deploy them.

Vonage goes vertical: healthcare, finance, retail

Also on June 3, Vonage (part of Ericsson) launched vertical-specific AI agents embedded directly in its Vonage Contact Center. Rather than building everything in-house, Vonage partnered with domain specialists: Avaamo for healthcare and Syndeo for financial services and retail.

The healthcare agents handle appointment scheduling, care navigation, billing support, and test results over voice channels. The financial services and retail agents blend deterministic logic with generative AI and flow-guided guardrails for secure, compliant automation. Both integrate into VCC without custom tool-switching, and both hand off to live agents with full context when needed.

This is a different bet from Meta's horizontal approach. Vonage is arguing that agents trained on industry-specific workflows, compliance requirements, and terminology outperform general-purpose alternatives. For a healthcare practice or a bank, that argument has merit.

Klaviyo: one config, 100+ languages

Klaviyo shipped a multilingual upgrade to its Customer Agent on June 3. Instead of duplicating agent configurations per market, brands configure once in English and the agent auto-detects and responds in over 100 languages across chat, email, SMS, and WhatsApp.

Multilingual support has been one of the most expensive parts of international expansion. Klaviyo's approach removes the need for separate agent configs per language, which cuts a real cost for e-commerce brands selling across borders.

MoEngage: MCP-connected marketing agents

MoEngage launched Merlin AI Custom Agents on June 3, giving lifecycle marketers and CRM teams the ability to build workflow agents on top of their existing data and tools. The agents connect through MCP (Model Context Protocol), include marketer-defined guardrails, and run within MoEngage's platform.

This is one of the first marketing automation platforms to ship MCP-connected agents with governance controls built in. For teams already running campaigns in MoEngage, it means agents can handle segment creation, flow building, and campaign optimization without leaving the platform or writing code.

Inline chart showing platform launches by category

What the convergence tells us

Five launches in 72 hours. Different company sizes, different channels, different verticals. The pattern:

Agents are the new default for customer-facing work. Meta is putting them in the hands of 200 million small businesses. Cisco is rebuilding enterprise contact centers around them. Vonage is embedding them in regulated industries.
The platform layer matters more than the model. Every one of these launches is about infrastructure — integrations, guardrails, governance, handoffs — not raw intelligence.
Vertical specialization is accelerating. Vonage's partnership approach (Avaamo for healthcare, Syndeo for finance) is built on the recognition that generic agents don't cut it in regulated industries.
Multilingual is no longer optional. Klaviyo's one-config, 100-language approach and Meta's local-language support both reflect a market where English-only is a growth ceiling.
Security and governance are selling points, not afterthoughts. Cisco's Agent 360, MoEngage's guardrails, Vonage's compliance framing — these are features, not disclaimers.

For businesses evaluating AI automation for intake, support, or lead qualification, the takeaway is straightforward: the tools exist now. The question isn't whether to deploy an agent. It's which platform's infrastructure and governance model fits your operation.

Sources:

Five Platforms, 72 Hours: AI Customer Support Agents Are Now Table Stakes