ESDS Software Solution is shifting from cloud services to artificial intelligence infrastructure. The company secured an agreement for 8,208 NVIDIA B300 graphics processing units to boost its business. As of September 2026, it operated five data centres. Investors now watch if this move will help the firm grow its revenue.

An artificial intelligence company can have the best engineers, the most advanced models and plenty of money. Yet none of that is of much use without the computing power needed to train and run those models. This is where a new category of technology companies is finding an opportunity. They provide the expensive infrastructure that lets AI companies do their work without building everything themselves.

For an Indian technology company, getting access to thousands of high-end graphics processing units (GPUs) and securing customers for that capacity is a significant opportunity. But there is a question that is just as important as getting the business. How much money is left after paying for the infrastructure?

This is the question investors need to ask about ESDS Software Solution. The company has moved from web hosting and cloud services into data centres, managed technology services and, more recently, artificial intelligence infrastructure. Its agreement involving 8,208 NVIDIA B300 graphics processing units (GPUs) could change its revenue profile. However, the economics of this business are not as straightforward as the headline contract value suggests.

ESDS Software Solution Share Price Chart

How ESDS Software Is Pivoting From Legacy Cloud to AI Infrastructure

ESDS began its business in 2005 as a web hosting and managed services provider. Over the years, it moved into data centres and cloud computing before developing its own technology platforms. Its patented Swaraj Cloud technology uses vertical auto-scaling, while its software portfolio covers cybersecurity, monitoring and automation.

The company has three broad business segments. Infrastructure as a Service (IaaS) provides cloud computing, data centre facilities and GPU infrastructure. Managed services cover the management and security of customers' technology systems, while Software as a Service (SaaS) includes the company's proprietary software products.

The idea is to sell more than one service to the same customer. A company may start by using ESDS for cloud infrastructure, then outsource its management and later use ESDS's software products. This can deepen customer relationships and increase revenue without requiring the company to find an entirely new customer for every additional service.

As of September 2026, ESDS operated five data centres. Its H1FY27 revenue retention rate was 86.7%, according to the company's investor presentation. These numbers provide a base for the next stage of growth, which is increasingly centred on artificial intelligence computing.

Inside ESDS's USD 1.25 Billion Sharon AI Deal for 8,208 NVIDIA B300 GPUs

The most visible part of this new strategy is an agreement with Australia-based Sharon AI. Under the arrangement, Sharon AI will deploy and operate a dedicated cluster of 8,208 NVIDIA B300 GPUs at an existing data centre facility in Australia. The agreement has a total contract value of USD 1.25 billion over five years, with an option to extend it by another two years.

ESDS has arranged for customers to use this computing capacity. The contracts are on a take-or-pay basis, which means customers pay for the available capacity irrespective of how much of it they actually use. This reduces the risk that ESDS will be left with expensive computing infrastructure and no corresponding revenue.

The five-year infrastructure agreement with Sharon AI has a stated contract value of USD 1.25 billion (around Rs 11,831 crore) and covers 8,208 NVIDIA B300 GPUs. Management expects the fully deployed cluster to generate monthly customer billings of around Rs 288 crore, implying an annualised run rate of Rs 3,456 crore. It has guided for a profit after tax (PAT) margin of 16-18%, translating into monthly PAT of around Rs 50-52 crore.

These are large numbers for a company whose FY26 revenue from operations was Rs 472 crore. The new project could therefore become much larger than the business ESDS has built over the past two decades.

ESDS has also paid Sharon AI an advance of USD 140 million, matched by a customer advance. The timing of amortisation and cash flows will be worth tracking as the project becomes operational.

Why ESDS Software's EBITDA Margin Fell to 41.8% Despite 47.5% Q2 Revenue Growth

The existing business has delivered healthy revenue growth. Revenue from operations increased 47.5% year-on-year in Q2FY27, while H1FY27 growth stood at 24.6%. For FY26, revenue growth was 31%, following 26% growth in FY25.

However, the operating margin trend is less impressive. The earnings before interest, tax, depreciation and amortisation (EBITDA) margin stood at 41.8% in H1FY27, against 46.4% in H1FY26. In Q2FY27, it was 41.8%, compared with 52% in the corresponding quarter last year.

Segment Breakdown: IaaS and SaaS Surge, While Managed Services Drops

This is worth examining because ESDS reported an operating margin of 50% in FY26, up from 43% in FY25. The latest performance suggests that investors should not assume that the higher margins will continue automatically as the business expands.

The segment numbers also reveal where growth is coming from. IaaS revenue increased around 52% in H1FY27, while SaaS revenue grew around 48%. Managed services revenue, on the other hand, declined around 10%. The company is becoming more dependent on infrastructure and software growth, while one of its established businesses has lost some momentum.

There is a difference between growing revenue and growing revenue profitably. As the AI infrastructure business becomes a larger part of the overall mix, investors will need to see whether its margins and cash generation can justify the capital being committed.

Customer Rationalisation: Why Account Count Dropped to 1,709 in Q2FY27

Another change is taking place in ESDS's customer strategy. ESDS wants to focus on its top 400-500 customers and move smaller accounts into a separate internal structure. The customer count fell from 2,216 in Q1FY27 to 1,709 in Q2FY27.

Larger accounts could generate more revenue and offer better profitability, but they also increase customer concentration risk. The 86.7% revenue retention rate in H1FY27 is another metric to watch as the company changes its customer mix.

Balance Sheet Check: Low 0.09x Debt vs. a 108% Jump in Other Liabilities

At first glance, ESDS does not have a large borrowing problem. Consolidated borrowings stood at Rs 113 crore as of September 2026, against Rs 96 crore at the end of FY26. The debt-to-equity ratio was low at around 0.09x and the company also had substantial cash and bank balances.

The September 2026 balance sheet, however, contains a less ordinary development. Other liabilities increased to Rs 2,702 crore from Rs 1,298 crore in March 2026. Other assets also increased to Rs 3,693 crore from Rs 1,541 crore over the same period.

Management explained part of this movement through the advance paid to Sharon AI and the corresponding customer advance received. This is valid, but investors should not treat every increase in liabilities as ordinary trade credit or every increase in assets as freely available resources. The timing and conditions attached to these advances matter.

ESDS raised Rs 720 crore through its September 2026 initial public offering (IPO). By September 30, it had utilised around Rs 46.9 crore, leaving approximately Rs 673.1 crore unutilised. Most of the unused money was parked in fixed deposits or the monitoring account, while Rs 576 crore had been earmarked for cloud computing equipment and data centre infrastructure.

The company therefore has money available for expansion and is not relying entirely on fresh borrowing. The question is how quickly this money gets deployed and what returns the new infrastructure generates.

Why Rising 81-Day Receivables Will Test ESDS's AI Cash Flows

Working capital deserves a closer look because a business can report strong revenue and still struggle to collect the money. ESDS reported days sales outstanding of 81 days in H1FY27, compared with 69 days in H1FY26. The annual debtor-days figure for FY26 was 128 days, against 145 days in FY25.

The collection period has improved on the annual comparison, but it remains long enough to require attention. This becomes more important when the company starts handling larger contracts, where payments to infrastructure partners and collections from customers may follow different schedules.

The operating cash flow number looks impressive, but the real test is whether the company can keep generating cash once the advances and initial funding arrangements have played out. Investors should watch collections, customer advances and payments to infrastructure partners together rather than looking at operating cash flow in isolation.

India AI Data Centre Expansion: Decoding the Rs 3,000 Crore Pipeline and 1 GW Target

ESDS is also planning to build AI infrastructure in India. Management said that it has an opportunity pipeline of around Rs 3,000 crore in the domestic market, largely for infrastructure requirements from government departments, defence public sector undertakings, banks and enterprises.

The company plans to invest around Rs 1,500 crore in the next phase of domestic capacity, using IPO proceeds and customer advances. It expects the first phase to become operational by June 2027, with around 1,150-1,200 GPUs. Management has indicated an annual revenue target of around Rs 1,000 crore from this business.

It is also looking to add more capacity during August and September 2027. Over time, management wants to establish more AI factories in India and overseas markets. It has discussed partnerships in the US, South Korea and Taiwan and disclosed a memorandum of understanding with a partner in Sweden.

Management aims to bring around 1 gigawatt of power under its control, focusing on locations with 50-100 megawatt capacity. However, the Rs 3,000 crore pipeline is not confirmed revenue, and the power target remains an ambition until backed by signed contracts.

Why Post-IPO Capital Dilution Dragged H1FY27 ROCE Down to 6.2%

ESDS has made progress in improving returns over the years. Its return on capital employed (ROCE) increased from 1% in FY23 to 12% in FY24, 21% in FY25 and 30% in FY26. Return on equity (ROE) also stood at around 25% in FY26.

ESDS reported a non-annualised return on capital employed (ROCE) of 6.2% and return on equity (ROE) of 6.6% in H1FY27, against 10.8% and 14%, respectively, in H1FY26. Part of the decline can be attributed to the Rs 720 crore IPO proceeds, most of which were yet to be deployed by September 2026, increasing the capital base without contributing fully to operating profits. However, the fall in EBITDA margin also suggests some pressure on operating profitability. The next few quarters will show whether deploying the fresh capital can improve these returns.

ESDS Software Share Price Valuation: Expensive at 103x P/E Even After a 39% Fall

At Rs 1,130 on November 20, 2026, the stock traded at approximately 103 times earnings. These are demanding valuations for a company whose largest new business is only now approaching the billing stage.

The stock has fallen around 39% from its 52-week high of Rs 1,864. A fall of this size is substantial, but it does not automatically make a stock cheap. The market may have reduced its expectations after the initial excitement around the listing, but the current valuation still assumes meaningful growth from the AI infrastructure business.

The Verdict: High-Stakes AI Execution Must Catch Up to ESDS's 103x Multiple

ESDS has an opportunity to build a much larger business than the one it started with. It has moved beyond web hosting, built cloud and data centre capabilities, developed proprietary software and secured a large AI infrastructure contract. Its domestic plans could provide another source of growth if management can convert the pipeline into funded projects and paying customers.

But the investment case should not rest on the number of GPUs or the total contract value alone. The Sharon AI economics need a clearer explanation, the updated billing timetable points to January 2027, and the company still has to show that the new infrastructure can generate cash and attractive returns.

The opportunity is real. So are the questions. At more than 100 times earnings, investors are paying for a future that has not yet fully arrived. The next few quarters will show whether ESDS is building a profitable AI infrastructure business or simply taking on a much bigger set of commitments in a fashionable market.

Note: We have relied on data from www.Screener.in throughout this article. Only in cases where the data was not available, have we used an alternate, but widely used and accepted source of information.

The purpose of this article is only to share interesting charts, data points and thought-provoking opinions. It is NOT a recommendation. If you wish to consider an investment, you are strongly advised to consult your advisor. This article is strictly for educative purposes only.

Manvi Aggarwal has been tracking the stock markets for nearly two decades. She spent about eight years as a financial analyst at a value-style fund, managing money for international investors. That's where she honed her expertise in deep-dive research, looking beyond the obvious to spot value where others didn't. Now, she brings that same sharp eye to uncovering overlooked and misunderstood investment opportunities in Indian equities. As a columnist for LiveMint and Equitymaster, she breaks down complex financial trends into actionable insights for investors.

Disclosure: The writer and her dependents do not hold the stocks discussed in this article. The website managers, its employee(s) and contributors/writers/authors of articles have or may have an outstanding buy or sell position or holding in the securities, options on securities or other related investments of issuers and/or companies discussed therein. The content of the articles and the interpretation of data are solely the personal views of the contributors/ writers/authors. Investors must make their own investment decisions based on their specific objectives, resources and only after consulting such independent advisors as may be necessary.