Cloud Computing

NxtGen Expands ‘M’ Open AI Platform with NVIDIA-Powered Inference to Enable Autonomous Enterprise Systems

NxtGen has announced the launch of India’s first sovereign, enterprise-grade agentic AI inferencing platform “M” built on NVIDIA Nemotron open models, enabling production-scale AI reasoning and autonomous decision-making across enterprise and industry workloads within India’s sovereign cloud framework.

Built on open-source foundations, the platform supports both global and India-developed AI models, enabling local-language and region-specific AI capabilities while ensuring full control over data, models and inference workflows. The platform is designed to operate entirely in sovereign environments with no dependency on external SaaS AI services.

M – India’s Sovereign Agentic AI Inferencing Platform

At the platform layer, M is NxtGen’s enterprise-grade, sovereign agentic AI inferencing platform optimised for NVIDIA Nemotron models and NVIDIA accelerated computing.

Unlike conventional AI copilots, M is designed as an intelligent inferencing layer that connects AI models with enterprise systems such as ERP, CRM, HRMS, ITSM and document repositories.

The platform is powered by:

● NVIDIA Nemotron LLMs and VLMs for reasoning and agentic tasks

● Open-source model frameworks to ensure transparency, extensibility and avoidance of vendor lock-in

● India-developed language models to support regional and multilingual AI use cases

● NVIDIA NeMo framework for model fine-tuning, deployment and runtime optimisation

● NVIDIA Nemotron Speech (ASR and TTS) for voice-first, multilingual interaction

● NVIDIA NIM microservices for GPU-optimised, low-latency inference

● NVIDIA Dynamo for efficient large-model reasoning using disaggregated prefill and decode

● An enterprise-grade AI stack optimised for scalability, security and deterministic inference

M enables enterprises to move from AI-assisted insight to agentic AI-driven outcomes through:

● Intent understanding and multi-step reasoning

● Secure tool invocation under governance guardrails

● Persistent state and memory across workflows

● Full observability, auditability and compliance

Organisations retain full control over models, data and inference logic, aligned with India’s data residency and regulatory requirements.

M for Travel: Vertical application on NVIDIA Nemotron

Built on top of “M” inferencing platform, M for Travel is one of the first vertical applications demonstrating how agentic AI can translate natural-language intent into structured, real-world journeys by reasoning across budgets, durations and constraints.

For example:

This intelligence layer allows travelers to express intent naturally - “Plan an efficient business trip with minimal layovers and a hotel near a quiet suburb close to New York’s World Trade Center”

M for Travel translates this intent into structured, executable journeys across planning, booking, and post-trip services, using live global travel inventory for flights, hotels and travel experiences through ITQ’s Travelport APIs, one of the world’s largest Global Distribution Systems (GDS). This provides the agentic AI layer with real-time access to airline, hotel and ancillary content, enabling the platform to move beyond itinerary generation to transaction-ready journeys spanning planning, booking and post-trip services.

India records more than 2.5 billion domestic airline trips annually and over 30 million outbound travellers, positioning the system requirements for high-concurrency, multimodal and persona-aware interaction at population scale.

At the infrastructure layer, NxtGen is building India’s dedicated AI Factory, powered by NVIDIA Blackwell GPUs, integrated with NVIDIA Spectrum-X Ethernet networking and NVIDIA BlueField-3 DPUs.

● This sovereign AI cluster supports large-scale model training and fine-tuning, high-performance inference, agentic and multimodal AI workloads, speech AI and AI-as-a-Service delivery across enterprises, startups, and government programs

While distinct from the inferencing platform, this cluster provides the compute foundation required to support population-scale AI workloads. The NVIDIA AI Factory is deployed entirely within India’s sovereign cloud framework, ensuring full data residency, regulatory compliance, and national-scale AI capacity for India’s growing AI ecosystem.

“NxtGen is building India’s AI enterprise inference layer,” said A. S. Rajgopal, MD & CEO, NxtGen. “With NVIDIA’s accelerated computing and NVIDIA Nemotron open models, we are delivering sovereign, production-grade AI systems that move beyond experimentation to autonomous, outcome-driven intelligence.”

Vishal Dhupar, Managing Director, Asia South, NVIDIA, said, “The future of enterprise AI in India is defined by the shift from AI-assisted insight to autonomous, agentic action. By deploying a sovereign, enterprise-grade platform powered by NVIDIA Nemotron open models k and NVIDIA Blackwell AI infrastructure, Indian businesses can operationalise intelligence with the data residency and control needed to fuel a new era of secure, outcome-driven AI at a national scale.”