Tech Mahindra
2025: A Transformational Year in Artificial Intelligence

Artificial Intelligence has emerged as the defining technology of the next phase of digital transformation. However, enterprises globally continue to face fragmented adoption, governance concerns, workforce preparedness gaps, dependence on external AI ecosystems and the absence of scalable enterprise platforms. Tech Mahindra recognised that realising the full value of AI would require a transformation that addressed technology, people, processes, governance and innovation simultaneously.

The company therefore launched an enterprise-wide AI strategy under the banner of “AI Delivered Right.” The strategy focused on four dimensions: Transformation Delivered, Productivity Delivered, Assurance Delivered and Innovation Delivered. It was aided by strategic interventions, including TechM Orion, a modular and centralis ed enterprise AI platform; VerifAI, an AI governance framework; Project Indus, an indigenous foundational language model initiative; and one of the largest workforce AI-skilling programmes in the industry. By 2025, these projects had enabled Tech Mahindra to establish itself as a recognised leader in enterprise AI while aiding to India’s broader ambition of becoming a global AI innovation hub.

Ihe rapid advancement of AI has created both unprecedented opportunities and significant challenges. One of the most pressing matters is the concentration of foundational AI capabilities within a limited number of global ecosystems, creating dependencies that can constrain innovation, linguistic inclusion and national technological self-reliance. India, with its linguistic diversity and scale, requires AI models that can serve local languages, dialects and contexts.

A second challenge relates to enterprise adoption. While organisations have invested heavily in digital transformation, many struggle to move from isolated AI pilots to enterprise-scale deployments that generate measurable business value. The absence of integrated AI platforms, governance mechanisms and operational frameworks often restricts scalability and increases risk.

The third challenge concerns workforce readiness. As AI technologies evolve rapidly, organisations require structured pathways to build skills, enable adoption and ensure employees can work effectively alongside intelligent systems.

Finally, the emergence of Agentic AI and autonomous enterprise models demands new frameworks for governance, transparency, accountability and responsible deployment. Dealing with these challenges required an integrated plan that combined indigenous innovation, enterprise enablement, workforce transformation and governance at scale.

The transformation remained anchored around a multi-layered AI ecosystem that combined foundational technologies, enterprise platforms, governance frameworks and workforce enablement to support the strategy.

At the foundation of this ecosystem was Project Indus, India’s first open-source Indic foundational large language model. Designed to support Hindi and more than 37 dialects, Project Indus was developed to democratise access to AI for over 1.4 billion people. Unlike conventional approaches that prioritise scale alone, the initiative adopted a frugal innovation model, beginning with a 1.2-billion-parameter architecture and scaling to an 8-billion-parameter model optimised for practical deployment.

Complementing Project Indus was TechM Orion, launched in July 2025 as a comprehensive platform for Agentic AI development. The platform provides capabilities for language model training, retrieval-augmented generation, autonomous agent development, orchestration and enterprise deployment. Equipped with more than 300 enterprise agents, Orion enables organisations to build and deploy AI solutions through automated development pipelines, modular architectures, multi-cloud support and integrated governance mechanisms, helping turn development capability into enterprise-scale deployment.

To ensure responsible deployment, Tech Mahindra introduced VerifAI, an AI governance framework that embeds transparency, accountability, fairness, explainability and compliance into enterprise AI systems. The framework reflects the organisation’s commitment to assurance as a core pillar of AI adoption and supports responsible deployment across enterprise use cases.

Recognising that technology transformation must be accompanied by workforce transformation, Tech Mahindra implemented a four-tier AI proficiency framework that created structured learning pathways across organisational levels. Supported by ecosystem partnerships with Microsoft, AWS, Google Cloud, Cognition and Cursor, the programme enabled approximately 77,000 employees to receive AI training, building the workforce foundation needed for enterprise-wide adoption and innovation.

The solution stack was further strengthened through Tech Mahindra’s strategic participation in the IndiaAI Mission, where it was selected as one of only eight organisations tasked with developing indigenous foundational AI models. This effort has positioned the company at the forefront of India’s AI sovereignty agenda and secured its role in building nationally relevant AI capabilities.

The transformation produced significant organisational, industry and national-level outcomes which strengthened its strategic direction.

Tech Mahindra successfully repositioned itself as an AI-first enterprise through the launch of the “AI Delivered Right” strategy and the creation of a comprehensive AI ecosystem spanning innovation, governance, workforce development and enterprise transformation. The company became the first systems integrator to launch a comprehensive platform for Agentic AI development through TechM Orion, which includes more than 300 enterprise agents and supports enterprise-scale AI deployment, turning platform capability into operational deployment.

Project Indus established a new benchmark for indigenous AI innovation by creating India’s first open-source Indic foundational LLM that supports 37+ dialects and is designed to serve 1.4 billion people. The programme also delivered an 8-billion-parameter education-focused LLM that was recognised among the top five LLMs at the IndiaAI Impact Summit 2026, linking model development to external recognition. This initiave is also aligned with NCERT curricula and designed for students in grades 6-12.

Workforce transformation was achieved at unprecedented scale, with approximately 77,000 employees trained in AI, creating a strong foundation for enterprise-wide adoption and innovation.

The initiative also received significant external validation. Tech Mahindra was selected as one of only eight organisations under IndiaAI Mission Cohort 2.0, recognised in Gartner’s Emerging Leaders category for Generative AI, included in the World Economic Forum’s MINDS Cohort, received the AVA Digital Award Gold Winner recognition and achieved finalist status at the Ragan Awards, reinforcing the initiative’s outcomes.

Beyond organisational outcomes, the initiative strengthened India’s AI ecosystem by advancing sovereign AI capabilities, promoting linguistic inclusivity, democratising access to AI technologies and supporting a nationally relevant AI innovation framework.

Tech Mahindra’s AI transformation demonstrates how enterprises can move beyond isolated AI initiatives to create a holistic ecosystem that integrates innovation, governance, workforce readiness and business transformation. Through Project Indus, TechM Orion, VerifAI, large-scale AI skilling and participation in the IndiaAI Mission, the organisation established a comprehensive framework for responsible and scalable AI adoption.

The initiative is significant not only for its technological achievements but also for its contribution to India’s AI sovereignty agenda. By supporting more than 37 dialects, serving the needs of 1.4 billion people, developing an 8-billion-parameter indigenous model, launching a platform with more than 300 enterprise agents. Tech Mahindra has created a replicable blueprint for AI-led transformation that combines business value with national impact.