Microsoft on Friday announced the statewide expansion of its AI-powered investigation platform for the Maharashtra Police. This will bring next-gen cybercrime tools to all 1,100 police stations across the state.
The rollout, unveiled on Friday during CEO Satya Nadella’s Mumbai leg of his India AI Tour, underscores Microsoft’s deepening partnerships with state governments, it said.
Maharashtra govt’s collab with Microsoft
The platform, known as MahaCrimeOS AI, was developed jointly with the Maharashtra Government and its MARVEL initiative, which was established to embed AI copilots across governance systems.
MahaCrimeOS AI runs on Microsoft Azure OpenAI Service and Microsoft Foundry, and is likely to provide a blueprint for another Indian states to strengthen their cyber-crime and digital safety systems.
Nadella on the AI wave
Addressing a packed auditorium at the Jio World Centre in BKC, Nadella said the world was entering a new era where the “frontier of what is possible” was rapidly expanding, driven by advances in artificial intelligence.
He said technology now allows governments, organisations and individuals to completely reimagine how they operate, from customer experience to employee productivity and overall operational efficiency.
Nadella noted that AI is reshaping innovation across sectors, including highly regulated industries such as pharmaceuticals.
He said drug companies can now explore whether they can accelerate clinical trials and bring more new medicines to market faster.
This, he said, reflects how industries are “moving on the frontier that we are all chasing.”
According to him, stepping into this new phase requires embracing fresh approaches rather than relying on methods that worked in the past. He stressed the need for both “unlearning and learning” as older patterns may not deliver results in the current technological landscape.
Explaining how AI development itself demands a new mindset, Nadella said building AI systems is very different from creating traditional information systems. Instead of starting with a detailed specification, AI development begins with defining the outcome. “You don’t start with a spec. You actually start with the test,” he said.
He added that developers now build rubrics or evaluations to measure the outcome, create a learning system around it and then continuously improve the model.
This process, he said, essentially inverts the traditional left-to-right approach of software building by beginning with the desired impact and working backwards.
Nadella emphasised that adapting to this new frontier means everyone must pick up new skills and use a modern tool chain every day. He said organisations need to bring together their data, tools and workflows to support these skills and mindsets, making AI a part of daily practice rather than an abstract concept.