Copyright yourstory
After the emergence of powerful Large Language Models (LLMs) like ChatGPT near the end of 2022, Sirish Kosaraju noticed that it was hard to adapt those LLMs for enterprise operations. To dig deeper, Kosaraju, former Co-founder and Director of RapidValue Solutions, joined hands with his colleagues at his former company, Rajesh Padinjaremada and Srinivas K. They eventually concluded that widespread adoption for enterprises would be challenging due to multiple issues, specifically concerning data security, governance, and the process of training LLMs with proprietary enterprise data. To fix that, the trio co-founded Wizr AI in 2023. Kosaraju took the role of its CEO while Padinjaremadam and Srinivas became its COO and VP. "What we are trying to do here with Wizr AI would be to extend the generic LLM capabilities and add a lot of features around enterprise-grade implementations. The architecture includes critical layers such as evaluations, observability, and versioning to smooth the transition from pilot projects to full production," Kosaraju tells YourStory. This platform ensures control over deployments and directly addresses privacy and security concerns by offering deployment flexibility. Kosaraju says it primarily operates in the US market. The Kochi-based startup has 54 employees and has onboarded 22 B2B clients across various industry verticals. Building the technology To achieve rapid implementation and value realisation, Wizr AI offers a no-code and low-code front end that allows users to create their own workflows and agents using a simple drag-and-drop interface. "What we really offer is essentially a platform to control all of these things from a centralised point of view rather than having different LLMs for different things. Tomorrow a company should be able to explain why an AI agent is making a decision. This means there is a legal implication and the company is liable for the decision. We give a complete end to end platform to be able to track all of these things," explains the CEO. Development started around the beginning of 2023 and the first release took approximately eight months. The process involved continuous feedback loops with enterprise partners, ensuring that each feature addressed genuine operational challenges. This collaborative approach enabled Wizr to fine-tune its platform's capabilities while maintaining alignment with evolving regulatory frameworks. A particular highlight of its tech is that it also provides deployment flexibility, including a bring-your-own-cloud model, which directly addresses crucial enterprise concerns about data security and privacy by allowing the platform to be hosted within the customer's platform (like an AWS). Kosaraju mentions that Wizr largely charges on two types of pricing: SaaS-based pricing (hosted on Wizr’s platform) and platform pricing (hosted on the customer’s platform). While the range varies depending on the amount of work, the average cost is somewhere between $150-200k per year. The deeptech startup competes with platforms like Writer AI and Glean. But Padinjaremadam says Wizr maintains an advantage because of its platform versatility, its ability to integrate with multiple enterprise systems and the flexibility to directly deploy on the enterprises' infrastructure. The market size for AI services is presently about $11.07 billion in 2025 and is projected to grow to $91 billion by 2035. Wizr AI is part of YourStory’s Tech30 cohort—a selection of India’s most promising startups of 2025—unveiled at TechSparks Bengaluru. (Edited by Affirunisa Kankudti)
