Air India has expanded Agentforce beyond its initial refund-automation use case to orchestrate three distinct service workflows: automated email resolution, name change processing, and knowledge assistance for frontline agents. The refund deployment already demonstrated material impact—reducing turnaround from 14 days to four hours—by handling multi-step validations, policy checks, and cross-system dependencies autonomously. The new deployments extend this pattern: email resolution now identifies multiple intents within single customer messages and routes them to specialised sub-agents with a 95% confidence threshold for autonomous response; name changes have dropped from three days to 30 minutes through end-to-end orchestration; and the knowledge agent consolidates access to policies, procedures and fare rules into a conversational interface for live-agent support. This progression reflects a deliberate shift from task automation to workflow redesign, underpinned by Air India's cloud-first architecture spanning 140+ systems.
The strategic implication is that Agentforce is moving beyond the narrow automation plays that characterised early AI agent deployments. Rather than replacing individual agent tasks, Air India is using it to eliminate handoffs and validation loops—the structural inefficiencies that compound resolution time. For CX teams already running Agentforce, this signals that the platform's value compounds when you move from isolated use cases to interconnected workflows; teams still piloting single-process automation may be leaving significant efficiency gains on the table. The 95% confidence threshold for autonomous response also matters: it suggests Salesforce is building guardrails that allow high-volume automation without sacrificing oversight, a critical distinction as teams scale beyond proof-of-concept.
Air India's broader context—30 agentic AI initiatives underway, evaluated against revenue, cost, experience and capability pillars—reveals how enterprise CX strategy is shifting. Automation is no longer a support function; it's becoming the primary mechanism for process redesign. The knowledge agent deployment particularly signals a move away from agent replacement toward agent augmentation, improving onboarding, first-contact resolution and consistency simultaneously. For support leaders evaluating AI investment, the question is not whether to automate, but whether your current platform architecture can orchestrate workflows across your existing system landscape without creating new integration debt.
Air India expands Agentforce to automate complex customer service workflows ET CIO