Enterprise AI is moving into an awkward middle ground: companies are using foundation models and AI agents at growing rates, but many still struggle to turn pilots into repeatable, governed production systems. Globant is betting that the next shift will involve changing not just how AI is built, but how enterprises buy it. The company has appointed Sarab Narang as CEO of Glob.AI, a new AI-native technology services model built around AI Pods, human oversight and pricing linked to outputs or consumption rather than hours or software seats.
Globant is positioning Glob.AI as more than another enterprise AI platform. The company is attempting to redesign the commercial model surrounding AI services, with autonomous and semi-autonomous agents performing defined work while human specialists remain responsible for supervision and governance.
The company announced the appointment of Sarab Narang as CEO of Glob.AI on August 6. Narang brings more than two decades of experience spanning generative AI, machine learning, enterprise software and consulting, including leadership roles at ServiceNow, Amazon Web Services and KPMG.
At the center of Glob.AI is the concept of an AI Pod. Globant describes each Pod as a service unit operated by a group of AI agents and supervised by human experts. Pods are intended to specialize in particular tasks or industries and can be accessed through an online, self-service model.
The distinction matters because Glob.AI is targeting the traditional technology-services model as much as it is targeting AI adoption.
Traditional consulting and IT services are commonly purchased around people, project scopes, hours or long-term contracts. SaaS companies, meanwhile, have historically monetized through users and seats. Glob.AI proposes tying its economics to the work performed or consumed instead.
That model arrives as agentic AI begins challenging assumptions embedded in enterprise software. Gartner estimates that up to $234 billion of enterprise application spending could be exposed to what it calls “agentic arbitrage” through 2030, as agents increasingly complete work across applications rather than requiring users to interact with each interface themselves.
In practical terms, the emerging architecture could look less like an employee opening five enterprise applications and more like an AI system coordinating work across those systems, with people supervising exceptions, approvals and higher-risk decisions.
Glob.AI is trying to package that transition as a service rather than asking every enterprise to build the entire stack internally.
The timing reflects a broader problem in enterprise AI. McKinsey’s 2025 State of AI research found that 88% of respondents said their organizations regularly used AI in at least one business function, yet roughly two-thirds had not begun scaling AI across the enterprise. Sixty-two percent said their organizations were at least experimenting with AI agents.
That gap between experimentation and scaled deployment creates an opening for technology-services companies. Enterprises may have access to models from Microsoft, Google, Amazon and other vendors, but deploying useful AI often requires workflow redesign, data integration, security controls, evaluation, governance and change management.
Globant’s pitch is effectively to bundle those capabilities into reusable AI-native service units.
The company’s early numbers suggest that it sees demand emerging. Glob.AI’s annual recurring revenue reportedly increased by roughly 60% in a single quarter as of June, while its pipeline reached $436 million. The company also says Glob.AI had adoption across 45% of Globant’s top 20 accounts. Those figures are company-reported rather than independently verified.
Narang’s background is particularly relevant to the strategy. At ServiceNow, he worked on agentic and generative AI products and go-to-market programs. Before that, he spent more than five years at AWS, ultimately becoming global head of generative AI and machine-learning go-to-market activities, with exposure to products including Amazon SageMaker and Amazon Bedrock. Earlier, he spent 12 years at KPMG working on AI, machine learning and technology engagements.
That experience places him at the intersection of AI infrastructure, enterprise software and services—the same boundaries that are increasingly blurring as AI agents become capable of completing multi-step business processes.
Glob.AI will nevertheless face substantial competition. Hyperscalers such as Amazon, Microsoft and Google are building agentic AI capabilities into cloud platforms and enterprise applications. Salesforce is embedding agents into customer workflows, while ServiceNow is integrating AI agents into its enterprise automation platform. Consulting and systems-integration firms are also developing AI engineering and managed-service offerings.
Globant’s differentiation is therefore less about owning a foundation model and more about packaging enterprise execution around models from multiple technology ecosystems.
The human-supervision component could prove equally important. Agentic systems create a new governance challenge because autonomy varies by task, access level and business risk. Gartner warned in May 2026 that 40% of enterprises could demote or decommission autonomous AI agents by 2027 because of governance failures.
For enterprise technology leaders, that makes the operating model as important as the AI model itself. A viable AI service must define who approves actions, how outputs are evaluated, what data agents can access, how failures are escalated and how performance is measured.
Glob.AI’s larger bet is that enterprises will increasingly purchase outcomes delivered by AI systems, rather than simply purchasing access to AI tools.
If that model gains traction, the implications extend beyond IT services. It could pressure SaaS vendors to rethink seat-based pricing, encourage consultancies to productize their expertise and create new categories of AI-native managed services.
The immediate test for Glob.AI will be whether AI Pods can deliver measurable business outcomes consistently across complex enterprise environments. The technology-services market has no shortage of AI pilots. What remains scarce is repeatable production value.
Market Landscape
The enterprise AI market is shifting from experimentation toward operationalization. McKinsey’s latest research shows that AI adoption is widespread, but only a minority of organizations have reached enterprise-scale deployment.
That creates three competing approaches.
Cloud platforms such as Amazon Web Services, Microsoft Azure and Google Cloud provide the infrastructure, models and development tools enterprises can use to build their own systems.
Enterprise software vendors such as Salesforce and ServiceNow are embedding agents directly into business applications, reducing the amount of custom infrastructure customers need to assemble.
AI-native services providers such as Glob.AI are taking a different route: delivering specialized AI capabilities as managed services, potentially allowing enterprises to purchase outcomes without building every agent, workflow and governance layer themselves.
Glob.AI’s output- or consumption-based pricing also places it within a larger debate about the economics of AI. If agents perform work that previously required employees to interact with multiple applications, traditional per-seat software economics could face increasing pressure. Gartner estimates that agentic arbitrage could put roughly 20% of enterprise SaaS spending at risk by 2030.
For CIOs and enterprise AI leaders, the relevant question is therefore not simply which model is most capable. It is which combination of models, agents, data, governance and human oversight produces the strongest measurable business outcome.
Top Insights
- Glob.AI introduces AI Pods combining AI agents and human supervision, giving enterprises a new way to purchase specialized AI services based on outcomes and consumption.
- Sarab Narang’s AWS, ServiceNow and KPMG background gives Glob.AI experience across enterprise AI products, machine learning infrastructure and technology-services commercialization.
- Glob.AI’s reported $436 million pipeline and adoption across 45% of Globant’s top 20 accounts indicate early enterprise interest, although the figures remain company-reported.
- Output-based pricing challenges traditional hourly consulting and seat-based SaaS economics as AI agents increasingly perform multi-step enterprise workflows across applications.
- Enterprise buyers will need strong governance, evaluation and human oversight as agentic AI moves from experimentation into production environments.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
