HYDERABAD: Two weeks after announcing its foundational artificial intelligence models for transcriptions, voice and images, Big Tech firm Microsoft wants to use its Azure cloud business strategy to sign long-term AI contracts with clients.
Microsoft’s business model for Azure has found success in the past 12 years of chief executive officer Satya Nadella’s tenure. The company, which follows a July-June financial cycle, reported $281.7 billion in revenue in FY25, out of which its cloud revenue hit $106.2 billion. Microsoft accounts for roughly one-fourth of the entire global cloud services industry’s annual revenue, which reached $419 billion in 2025, as per market research firm Synergy Group.
In an interview with Mint, Jay Parikh, executive vice-president and member of Microsoft’s global leadership team, said the company will seek to use the same playbook for AI as well.
“We currently have 11,000 models and 80,000 customers of it in our ‘foundry’ and this number changes very rapidly. Our go-to market strategy is to offer our customers a fully flexible AI platform where clients can use a toggle to choose the safest AI model that might be slower in performance or the most cutting-edge ‘frontier’ AI models too. This includes our own AI models as well that we just introduced,” Parikh said.
Microsoft has previously offered its ‘365’ suite of applications in cloud services and has given businesses the flexibility to use their preferred third-party applications such as Slack by Salesforce and SAP’s enterprise resource planning platform. Parikh said the company aims to do the same with AI as well—for both models and applications.
On the sidelines of his visit, Microsoft hosted an internal hackathon, ‘Pinnacle’, at its Hyderabad engineering campus. On Thursday, almost 500 engineers put up over 100 booths to showcase internal applications and services to Parikh, who also heads the company’s ‘CoreAI’ division that built its own AI models.
Filling strategic gaps
The idea, Parikh said, “is to try and understand how we can use and leverage AI to fill in those key strategic gaps that keep a company away from converting a pilot project into a full-scale AI deployment.”
Two in-progress projects that Microsoft’s employees showcased, which Mint was privy to, were Microsoft’s version of Anthropic’s ‘Cowork’ platform that runs third-party apps inside its own interface, and an ‘approvals’ mechanism for managerial executives to autonomously read information from Slack, Gmail and WhatsApp. Both these features are in the early stages but are expected to be a part of Microsoft’s ‘Copilot’ AI assistant platform in the coming quarters.
The tech giant’s Hyderabad campus, spread across 3 million square feet, houses about 10,000 engineers and is Microsoft’s largest presence outside of its Redmond, Washington headquarters in the US. The company also houses engineers across Bengaluru, Noida and Pune.
While Parikh did not underline specific applications that were built solely out of its India offices, he said that on GitHub, the Microsoft-owned code repository platform, India currently accounted for “27 million out of the world’s 180 million developers, and also the single-largest country with the most open-source software developers.”
India, the top executive added, is also important for its IT services industry, which Parikh believed would remain resilient even as the impact of AI reshapes it.
“A lot of our services are actually built after consultation with… TCS, Infosys and the likes. We have decades of partnerships with them, and they have the closest view of real-world gaps and problems that clients across energy, financial services and other industries have. Our partnerships with tech service integrators help us understand these and then engineer applications and AI models in line with what these gaps are, and I don’t see this changing,” he said.
Analysts said Microsoft’s flexibility approach could work well, with the company leading OpenAI and Anthropic in revenue.
Flexibility is key
“Azure’s business model has always been based on flexibility and has done well for Microsoft. In the long run too, the company that wins the enterprise AI race will be the one that gives the most amount of choice,” said Kashyap Kompella, founder and chief of technology consultancy firm, RPA2AI.
Microsoft’s longstanding government relations mean that it has a certain degree of trust at policy levels that companies would any day prefer over a startup, he said.
“While Amazon is focusing its AI efforts within itself and Google so far taking a consumer-centric approach as it had with cloud, Microsoft is widely regarded as being in an advantageous standing in AI,” Kompella said.
Through the four quarters of calendar 2025, Microsoft reported $120.5 billion in revenue from ‘Intelligent Cloud’, which includes its Azure cloud services as well as AI products and solutions. Microsoft does not break down the divide between cloud and AI revenue.
In comparison, Anthropic hit an annual revenue run rate of $9 billion in calendar 2025 and expects to earn $30 billion in this calendar year. On 19 June, OpenAI’s chief financial officer Sarah Friar said the company’s revenue touched $20 billion in calendar 2025.