Fractal Analytics Ltd

The board of Fractal Analytics at its meeting held on 20 May 2026 has approved investment of USD 0.6 crore (INR equivalent of 57.6 crore), by Fractal US, wholly owned material subsidiary of the company, in Asper US, step down subsidiary of the company.

The company stated that in Q4 FY26, growth was led by a strong performance in the healthcare and life sciences (HLS) segment which grew 82% YoY and the banking, financial services, and insurance (BFSI) segment which grew 42% YoY. The consumer-packaged goods and retail (CPGR) segment grew at 11% YoY, while the telecom, media and technology (TMT) segment declined 19% YoY due to client-specific issues. Fractal further expanded its relationships with existing clients resulting in a net revenue retention (NRR) of 112% in Q4 FY26. Total operating expenditure for the period under review was Rs 719.60 crore, up 13.3% YoY. Fractal's gross margin in Q4 was 48.2%, an expansion of 47 basis points YoY. Adjusted EBITDA margin expanded by 189 basis points to 22%. Profit before tax in Q4 FY26 stood at Rs 143.10 crore, up by 66.6% from Rs 85.90 crore in Q4 FY25. For FY26, the company has recorded net profit and revenue from operations of Rs 292.20 crore (up 31.0% YoY) and Rs 3299.70 crore (up 19.3% YoY), respectively. Srikanth Velamakanni, Group CEO said: 'We wrapped up FY 2026 on a strong note, with robust revenue and profit growth while delivering AI-led transformation for our clients. AI is becoming more capable everyday: AI that can plan, reason, and act through complex enterprise work - and this frontier intelligence is becoming much more affordable to deploy. This, in no uncertain terms, means that enterprise AI is taking off. And it is exactly what Fractal was built for.' Fractal is a publicly listed global enterprise AI company with a vision to power every human decision in the enterprise. The scrip rose 0.13% to currently trade at Rs 1,071 on the NSE. Powered by Capital Market - Live

Net profit of Fractal Analytics rose 114.96% to Rs 117.80 crore in the quarter ended March 2026 as against Rs 54.80 crore during the previous quarter ended March 2025. Sales rose 17.00% to Rs 886.30 crore in the quarter ended March 2026 as against Rs 757.50 crore during the previous quarter ended March 2025. For the full year,net profit rose 31.03% to Rs 292.20 crore in the year ended March 2026 as against Rs 223.00 crore during the previous year ended March 2025. Sales rose 19.32% to Rs 3299.70 crore in the year ended March 2026 as against Rs 2765.40 crore during the previous year ended March 2025. ParticularsQuarter EndedYear EndedMar. 2026Mar. 2025% Var.Mar. 2026Mar. 2025% Var. Sales886.30757.50 17 3299.702765.40 19 OPM %19.5216.17 -13.2711.58 - PBDT186.00116.50 60 453.50313.30 45 PBT149.4082.60 81 317.70211.00 51 NP117.8054.80 115 292.20223.00 31 Powered by Capital Market - Live

Fractal (www.fractal.ai) has secured a CDP 'B' rating for the third consecutive year, underscoring robust climate action governance and emissions management in 2025. CDP, recognized globally as a leading environmental emissions reporting platform, assessed more than 24,800 companies worldwide in 2024, underscoring the scale and scope of its benchmarking process. By measuring and disclosing greenhouse gas emissions, companies contribute to the comprehensive datasets on corporate climate action, enabling investors, policymakers, and other stakeholders to evaluate progress, risks, and opportunities. CDP scores also help organizations identify strengths and gaps in environmental management, guiding sustainability strategies. A CDP rating of 'B' reflects a coordinated management of climate-related impact, risks, and opportunities, along with providing demonstrated evidence of actions associated with good environmental management.Powered by Capital Market - Live

The board of Fractal Analytics at its meeting held on 20 March 2026 has approved the investment of USD 0.3 crore, by Fractal Analytics Inc. (Fractal US), wholly owned material subsidiary of the Company, in Asper.Ai Inc. (Asper US), step down subsidiary of the Company. Powered by Capital Market - Live

The board of Fractal Analytics at its meeting held 20 March 2026 has approved an increase in the investment limit in Fractal US, a wholly owned material subsidiary of the Company, by an INR equivalent of USD 4.5 crore, to invest the funds in one or more tranches, such that all investments by the Company in Fractal US including all previous investments shall not exceed INR equivalent of USD 15 crore. Out of the proposed investment of USD 4.5 crore, INR 262.3 crore (approx. USD 2.8 crore) will be utilized for repayment of loans availed by Fractal US, which forms part of the use of proceeds towards one of the objects of the Offer (IPO), and the balance will be utilized strategically for working capital requirements and strengthening the capital base of Fractal US. Powered by Capital Market - Live

The board of Fractal Analytics at its meeting held 20 March 2026 has approved the investment of Rs 39.4 crore in one or more tranches by way of subscription in equity shares of Analytics Vidhya Educon (AVEPL), a wholly owned subsidiary of the Company. AVEPL was incorporated on 18 February 2014. It is engaged in the business of providing a web platform service to impart knowledge related to analytics through paper and web articles, competitions, community events, and also includes corporate training, placement, and job provider services. Powered by Capital Market - Live

The board of Fractal Analytics at its meeting held 20 March 2026 has approved the sale and transfer of the business units namely EdTech, Fractal Analytics Academy (FAA) and Iqigai.ai together with their respective employees, contracts, consultants, intellectual property rights, licenses and such other assets and liabilities as may be identified by the Company and attributable to the said business units, to Analytics Vidhya Educon (AVEPL), a wholly owned subsidiary of the Company, as a whole and on a going concern basis, by way of a slump sale, for an aggregate consideration of Rs 10.9 crore. Powered by Capital Market - Live

The company said LLM Studio is designed to help enterprises build, customise, deploy and manage domain-specific language models using open-source frameworks and NVIDIA's AI infrastructure. The platform will be showcased at NVIDIA GTC 2026, a key global event for AI and accelerated computing. At its core, LLM Studio brings together two modules. AutoLLM enables businesses to create smaller, task-specific models with capabilities such as model selection, synthetic data generation and benchmarking. LLMOps, on the other hand, focuses on deployment, monitoring and governance across the model lifecycle. Fractal highlighted that enterprises are increasingly moving beyond one-size-fits-all large language models and are seeking solutions that offer better control, cost predictability and reliability in production. The platform aims to address these needs by allowing organisations to build proprietary models grounded in their own data, thereby reducing hallucinations and improving output accuracy. Built on NVIDIA's ecosystem, including NeMo for model development and NIM microservices for deployment, the platform is designed to standardise AI operations across cloud environments while reducing the need for custom builds. Fractal Analytics, a global enterprise AI company, serves Fortune 500 clients and continues to expand its capabilities in AI-driven decision-making, with a growing focus on production-ready, enterprise-scale applications. The company reported 10.6% jump in consolidated net profit to Rs 102.60 crore on 20.8% increase in net sales 854.40 crore in Q3 FY26 over Q3 FY25. Powered by Capital Market - Live

Fractal Analytics announced the launch of LLM Studio, an enterprise platform that helps organizations build and run language models tailored to their business. It is designed for teams that want more control over how models are governed, deployed, and managed in production. LLM Studio will be demonstrated at NVIDIA GTC 2026, the premier AI and accelerated computing conference by NVIDIA, taking place from March 16-19, 2026, at the San Jose McEnery Convention Center in California. Enterprises are moving beyond one-size-fits-all, API-only large language models for high-value use cases. Today, business and technology leaders need clearer guardrails for governance, predictable costs, and reliable performance. At the same time, industry research points to growing adoption of smaller, purpose-built models that can be tuned to specific functions and domains. LLM Studio enables businesses to design, build, evaluate, and operate domain-adapted language models using open-source models, powered by NVIDIA AI infrastructure and software stack. It brings together two modules: AutoLLM: Helps businesses create smaller, specialized models for specific tasks or industries. It supports open-source model selection, synthetic data generation, model customization, evaluation, and performance benchmarking. LLMOps: Helps teams manage the full life cycle after a model is created. It supports deployment, monitoring, and governance. LLM Studio helps keep model responses tied to an organization's approved data and context. This reduces hallucinations and improves the quality of reasoning. The resulting models remain proprietary to the organization. Teams can then use these models in agents or other generative AI applications to deliver more reliable performance, often at a fraction of the cost of running larger foundation models.Powered by Capital Market - Live