“R&D India HQ, local engineering team, partner expansion, and IPO roadmap are on our agenda,” say Altos CEO and India Director

The company sharpens India strategy around full-stack AI infrastructure as enterprises move from pilots to production deployments.

(Left: Jackie Lee, CEO, Altos Computing; Right: Priya Krishnamurthy, Director, Altos India)

Altos Computing, an Acer group company, is aligning its India strategy with the next phase of enterprise AI adoption, as organisations shift from pilots to production deployments. The company has rolled out a Make-in-India AI server portfolio, including the Altos BrainSphere R300 AI Series, targeting enterprise, government, research and hyperscale workloads.

Speaking to CRN India, Jackie Lee, CEO, Altos Computing and Priya Krishnamurthy, Director, Altos India, said the company is moving beyond hardware to deliver integrated AI infrastructure, built on an EMS-led model and bundled with its software stack for deployment-ready systems.

“Currently, we are working through a network of EMS partners. At this stage, we are not involved in component manufacturing. Instead, we have identified vendors who provide various subsystems, which are assembled at our facility and bundled with our software,” Krishnamurthy said.

The company works with multiple EMS vendors across sub-components, bringing them together into a unified system for deployment at customer sites.

“We define specifications based on customer workloads, including performance and hardware requirements, identify and qualify local partners, and carry out final assembly at our facility in Puducherry,” Lee said.

He added that Altos integrates its software with locally assembled hardware to deliver end-to-end AI infrastructure for both development and production workloads.

As this is the first product line, the company plans to expand its partner ecosystem in line with rising demand, with more EMS partners expected to be added over time.

Outlines India IPO roadmap, local R&D and engineering push

Altos is laying out a long-term roadmap for India that includes plans beyond its current operational footprint.

The company has a roadmap to pursue an IPO in India alongside its global plans.

Krishnamurthy said, “There is a clear roadmap from Lee to pursue an IPO for Altos in Taiwan, along with a roadmap for an IPO in India. Such plans reflect a strong long-term commitment to the market and the business.”

At the same time, Altos is working towards establishing an R&D presence in India and building a local engineering team.

Currently, most of the engineering capability is based in Taiwan, but the company sees a need to develop in-country expertise as demand scales.

“Why not have a local engineering team who can create from the silicon level itself,” Krishnamurthy said, pointing to the growing domestic demand for AI infrastructure.

Lee added that the next phase will also include investment in software R&D headquarters in India, particularly to support enterprise customers with compatibility and performance optimisation.

Customers today are looking for solutions rather than standalone products, he said, highlighting the need for localised engineering and R&D capabilities.

From servers to full-stack AI infrastructure

Altos wants to be more than a hardware provider, focusing on delivering a complete AI infrastructure stack that integrates hardware, software, and ecosystem components.

After the final assembly in Puducherry, the systems are integrated with Altos’ in-house software.

“We are not only offering hardware,” Lee said.

Altos integrates its software stack, including tools designed for AI computing and workload management, with infrastructure to deliver deployment-ready systems. These solutions are aimed at both AI development and production workloads.

Krishnamurthy said the company’s focus is on creating a plug-and-play ecosystem for customers. This includes integrating storage, networking, and other components to deliver a complete environment that is ready for deployment.

A key area of focus is data sovereignty. Organisations are increasingly looking to retain control over their data, particularly in sectors including government and healthcare.

Another key aspect, according to Krishnamurthy, is addressing the challenge of GPU utilisation, as enterprises continue to invest in AI infrastructure.

Customers procure GPUs without fully understanding how effectively they will be used, she said.

“Do not just secure the GPU… use it more wisely,” she added.

To address this, Altos bundles software with its servers to enable better utilisation of GPU resources. This includes capabilities, including virtualisation, where multiple workloads can run on the same GPU infrastructure, and analytics that help administrators track usage patterns.

The system enables multiple instances and workloads to run simultaneously, allowing organisations to optimise resource allocation.

The software is provided as a perpetual offering along with the server, rather than as a subscription, helping reduce the overall cost of ownership.

Partners move from resellers to lifecycle owners

Altos is expanding its partner ecosystem as AI adoption shifts towards production, requiring deeper engagement across the deployment lifecycle.

The company is working with both system integrators and independent software vendors (ISVs) to combine infrastructure and application capabilities.

“We are expanding our channel network,” Lee said, noting that AI customers often lack in-house capabilities to build and deploy solutions.

Partners are expected to play a role across the entire lifecycle—from identifying customer requirements and providing financing models to deployment, optimisation, and even recycling of infrastructure.

Krishnamurthy said the company is also exploring OPEX-based models to make adoption easier for customers, particularly SMEs.

The lifecycle approach extends to sustainability as well, with a focus on responsible recycling of products at the end of their lifecycle.

“We want a customer to start the journey cycle and end it in a very responsible manner,” she said.

Building CoEs and addressing operating gaps

Altos is also investing in centres of excellence (CoEs) to support customer adoption.

The company’s upcoming headquarters in Bangalore will include a CoE where customers can test and run applications on live infrastructure. This goes beyond traditional demos, allowing customers to access systems remotely and evaluate performance.

At the same time, Altos is enabling partners to set up their own CoEs across cities, supported through demo programmes and proof-of-concept environments.

“We will have a mix of both,” Krishnamurthy said, referring to vendor-led and partner-led CoEs.

Despite these investments, the company sees a gap in operating AI environments.

“Operating is the challenge,” she said, pointing to issues such as prompting and effective utilisation of AI tools.

The ability to ask the right questions and use AI systems effectively remains a key learning curve for customers, she added.

Demand rises across sectors as AI moves to production

Altos sees long-term demand for AI infrastructure across multiple sectors, including government, defence, education, and SMEs.

Government and defence are expected to be significant drivers, particularly for on-premises and air-gapped deployments.

Education continues to be a steady market, while SMEs represent a growing opportunity as businesses adopt AI technologies.

“Even a simple agent chatbot will get adopted,” Krishnamurthy said.

Lee noted that AI adoption has accelerated following the rise of ChatGPT, which demonstrated the potential of generative AI to improve productivity.

Earlier, between 2018 and 2020, AI adoption was limited to specific use cases such as security, computer vision, and healthcare. However, the lack of a clear ROI led some organisations to step back.

This has now changed, with enterprises recognising the value of AI and moving towards production deployments.

“The demand for AI will grow very fast,” Lee said.

Partners must shift to consultative engagement

As AI adoption evolves, Altos is also redefining what it expects from partners.

Partners need to move beyond product-led selling and focus on understanding customer problems.

“Partners should begin by clearly understanding the customer’s problem statement. The focus should not be on selling a server, but on identifying and solving the customer’s specific needs,” Krishnamurthy said.

She added that if partners can address the customer’s problem effectively, it builds long-term relationships.

The approach should be consultative rather than purely sales-driven, as solving the right problem helps in retaining customers over the long term.

Lee added that partners must be able to translate customer demand into clearly defined workloads, which can then be supported by Altos’ infrastructure.

This shift towards consultative engagement is critical as AI deployments become more complex and require alignment between infrastructure and application requirements.