NxtGen has introduced M, India-built platform uses multi-model orchestration to turn intent into action across enterprise services.
NxtGen has launched M, an open-source GenAI platform designed to transform human requests into executable actions through intelligent, agentic workflows. Built in India, M aims to bridge advanced AI capabilities with practical, everyday use cases while preserving data sovereignty.
The platform acts as an orchestration layer for thousands of open-source large language models. It dynamically selects the best model for each task, from domain-specific datasets for finance or law to expert reasoning models like DeepSeek 671B. Inputs can be text or voice, with M integrating via APIs to external services. Whether it is for booking tickets, processing documents, or provisioning enterprise infrastructure. This multi-model approach delivers context-aware responses without token restrictions. Enabling deep reasoning over long interactions.
M addresses key AI adoption challenges in India: limited infrastructure, concentration of AI innovation in a few hands, and lack of tools for startups and SMEs. By running entirely on NxtGen’s infrastructure with controlled internet access, M ensures security and compliance for regulated sectors. Live demonstrations showed it building a website from scratch, deploying virtual machines on NxtGen’s Speed Cloud, and extracting insights from complex documents—all without human coding.
Enterprises can embed their services into M for automated workflows, while public-facing integrations could simplify citizen services like voter ID applications or train bookings. Positioned as a sovereign-ready alternative to closed models, M can swap in newer, better-performing models without lock-in. NxtGen plans to expand industry-specific capabilities and add autonomous task execution. CEO Rajgopal A.S. says the mission is clear: “We want AI to work for everyone, seamlessly, securely, and at scale.”
