Adobe India Says Firms Fix Workflows Before Deploying Agentic AI

Prativa Mohapatra, vice president and managing director at Adobe India, told the Economic Times that nearly a third of Adobe's global innovation, including Firefly AI, is driven by teams in India, and that Adobe has around 8,000 employees in the country. Speaking to ET, Mohapatra said enterprises are "increasingly working on change management solutions before deploying agentic AI tools." She added, "How you are going to implement (agentic AI) solutions that nobody has ever done before," and called the shift "both a challenge and a great opportunity for system integrators. So, they are partnering with us." (Economic Times)

Companies adopting agentic AI commonly face nontrivial integration work that spans orchestration, data pipelines, and operator interfaces. Observer experience in the field shows that moving from a generative model prototype to a production agent often requires: robust input/output schemas, centralized observability, incremental rollout mechanics, and layered guardrails for safety and compliance. These technical needs make predeployment workflow redesign a recurring, engineering-heavy task rather than a purely product or UX change.

Industry reporting frames Adobe's comments as part of broader enterprise adoption patterns where AI agents shift the locus of value toward end-to-end processes. For practitioners, this trend elevates cross-functional requirements: platform engineers must collaborate with product, security, and compliance teams early in the lifecycle. System integrators and consulting partners appear in public coverage as common intermediaries in larger deployments, reflecting demand for integration expertise across sectors such as fashion, jewellery, travel, and financial services (Economic Times).

Observers and practitioners should track three signals:

• •how vendors expose orchestration and observability primitives in commercial agent stacks
• •case studies showing phased rollouts or change-management programs tied to agent deployments
• •the emergence of standardized guardrail tooling for agentic workflows. These indicators will show whether the current emphasis on predeployment workflow fixes produces repeatable integration patterns or continues to rely on bespoke system-integration work