The evolving landscape of enterprise operations is increasingly defined by the integration of generative artificial intelligence (AI). Amidst this transformation, Genpact is carving out a distinct path characterised by its ambitious ‘Agentic Enterprise’ model, which posits that AI not only supports business processes but fundamentally becomes the processes themselves.
Aniruddha Ray, Senior Vice President of Agentic Products and Platform Engineering at Genpact, recently articulated this transformative vision during the AIM DES 2025 event. He described the ‘Agentic Enterprise’ as “the most powerful construct in my lifetime,” highlighting a significant shift from traditional operational models. Under this new framework, Genpact’s transformation strategy, known as the ‘Genpact Next Strategy’, is anchored in three pillars: advanced technology, prioritising service lines, and unlocking value for customers.
Ray differentiated between agent-powered and agentic enterprises, emphasising that in agent-powered scenarios, AI acts as a complement to existing workflows. In stark contrast, agentic enterprises see AI at the heart of the process. He explained, “Today, if I run a process and I embed AI or software in it, what will happen is that the process will be driven by AI, and humans will actually enable part of it.” This paradigm shift underscores the evolving interplay between humans and AI, where human roles transition from active participants to strategic overseers.
To facilitate this transition, Genpact is championing advanced data strategies. Traditional static and siloed data is being replaced by unified, contextual, and real-time data streams powered by knowledge graphs and semantic ontologies. Ray elaborated on this evolution, noting that complex application interfaces are gradually giving way to more intuitive, agent-driven interactions.
A crucial innovation in Genpact’s endeavour to anchor its agentic approach is the introduction of ‘systems of innovation’. This orchestrated layer allows agents to interact autonomously with each other, as well as with both humans and systems. Ray identified this as the essential component missing from previous models and pivotal for achieving true agentic functionality.
Central to these developments is Genpact’s initiative to “supercharge knowledge workers” with cutting-edge tools such as Dataverse and Cora Code GenY, which facilitate rapid development of industry-specific models and software solutions, respectively. The company’s recently launched AI GigaFactory stands as a testament to this innovative drive. This large-scale automation framework allows for accelerated deployment of specific use cases, potentially reducing project timelines from six months to as little as two to four months, according to Ray.
Genpact’s approach to success has evolved as well, now prioritising tangible business outcomes over mere automation metrics. For instance, the company reports that it has reduced costs related to invoice processing by 70% and significantly improved claims handling speed. This shift towards measuring success through impact rather than automation percentage underscores a deeper commitment to delivering real value to clients.
Two major operational shifts have been introduced within Genpact—’service-as-agentic-solutions’ and the AI GigaFactory model. The former represents a transition to solutions driven by software, while the latter embodies an innovative delivery model aiming to drive impactful AI integration. Together, these shifts aim to create a tightly-knit architecture connecting systems of record, insight, and engagement, powered by a specialised agentic AI engine.
However, the path to achieving these ambitious goals is not without challenge. Ray identified three critical obstacles that Genpact faces: ensuring data quality, achieving seamless integration, and selecting the right agents for specific tasks. He noted, “We are realising we need to create an ensemble of agents… determined by the type of data it’s applying to.” This meticulous selection process is underscored by an emphasis on domain expertise and semantic understanding.
As Genpact continues to implement its vision across various sectors—including finance and risk, accounts payable, and other operational areas—the company is keenly aware of the necessity for robust capability-building models. This effort begins with a comprehensive data engine room that prepares anonymised or synthetic data for agent training. Central to this is the AI agent foundry, which aims to refine multiple agents configured with domain-specific large and small language models (LLMs and SLMs).
Ray emphasised the importance of collaboration in this endeavour, stating, “You don’t do anything without working with partners.” The strategy includes not only direct partnerships but also leveraging agent marketplaces and proprietary tools to enhance Genpact’s capability.
With the rollout of its G Solutions marketplace, featuring over 1,000 active AI models deployed across its platforms, Genpact is positioning itself as a leader in the agentic landscape. Although not all its models are designed to be agentic, the company has already developed 21 executable agents, with plans to scale this count to 50 by year-end. The anticipated productivity gains of up to 40% from recent AI implementations demonstrate a promising trajectory, with aspirations for a further 60–70% boost over the next two years.
In this fast-evolving technological era, Genpact’s commitment to transforming enterprise operations through agentic methodologies reflects a broader trend towards the integration of advanced AI systems. By prioritising innovation, strategic partnerships, and a clear focus on value creation, Genpact is not only redefining its operations but also setting benchmarks for others in the industry.
Reference Map
1: Paragraphs 1, 2, 3, 4, 5, 6
2: Paragraphs 4, 7
3: Paragraph 8
4: Paragraph 7
5: Paragraph 8
6: Paragraph 1
7: Paragraph 1
Source: Noah Wire Services