Deccan AI Raises $25M in Series A Funding Round

Deccan AI Secures $25 Million in Series A Funding Led by A91 Partners, Prosus, and SIG

The Indian AI funding ecosystem received a significant vote of confidence as Deccan AI, a fast-rising startup specialising in post-training data and model evaluation services, successfully closed a $25 million Series A round. The all-equity funding was spearheaded by growth equity firm A91 Partners, with participation from Susquehanna International Group (SIG) and Prosus Ventures. This latest AI funding news reflects the growing global appetite for specialised AI infrastructure companies that go beyond model building and focus on making AI systems production-ready, accurate, and enterprise-safe.

For the broader AI community, this investment signals more than just capital injection into a single startup. It points to a structural shift in how the AI industry is evolving — one where post-training services, model evaluation, and enterprise-grade deployment tools are emerging as critical and high-value components of the AI supply chain. Deccan AI's raise is being closely watched by investors, technologists, and enterprises alike, as it validates a growing niche that sits at the intersection of data engineering, reinforcement learning, and real-world AI deployment. The AI World Organisation, which has consistently tracked global AI funding trends, recognises this milestone as a landmark development in the rapidly maturing AI funding landscape.

From Frontier Labs to Enterprise Powerhouse

Deccan AI was founded in 2023 by Rukesh Reddy, a visionary entrepreneur who identified a critical gap in the AI development pipeline — the post-training phase. While most of the public discourse around artificial intelligence focuses on the development of large language models and foundation models by companies like OpenAI, Anthropic, and Google DeepMind, the phase that follows initial training is often overlooked yet arguably just as important. Post-training involves refining AI models through techniques such as reinforcement learning from human feedback (RLHF), domain-specific fine-tuning, nuanced evaluation, and ensuring that these systems behave reliably when deployed in messy, real-world business environments.

In its early days, Deccan AI concentrated on providing high-quality data and model-training support to some of the most advanced AI research labs in the world. The company built deep expertise in curating complex, domain-specific datasets that frontier labs needed to move their models from decent to exceptional. This foundation gave Deccan AI a unique vantage point — it understood, better than most, where AI systems fail, why they fail, and what kind of data and evaluation infrastructure is needed to fix those failures at scale. Over time, this insight shaped the company's strategic pivot toward enterprise solutions, transforming it from a services provider for research labs into a full-stack platform company serving large global businesses.

Today, Deccan AI operates from its headquarters in the San Francisco Bay Area, with a substantial operations team based in Hyderabad, India. The company employs around 125 full-time professionals and complements this core workforce with a remarkable contributor network of over one million individuals — including university students, domain specialists, and PhD researchers. In any given month, between 5,000 and 10,000 of these contributors are actively engaged in projects, making Deccan AI's talent pipeline one of the most expansive in the AI data services space. With a new Bengaluru office now in the works, the company is doubling down on its India presence with a specific focus on enterprise growth across the Asia-Pacific region.

Three Products Powering the Future of Enterprise AI

One of the most compelling aspects of Deccan AI's story is the depth and specificity of its product portfolio. Rather than offering broad, generic AI tools, the company has built a tightly focused suite of three products designed to address the most pressing challenges enterprises face when trying to operationalise AI at scale. These products reflect a deep understanding of what large organisations actually need — not just cutting-edge models, but dependable, auditable, and continuously improving AI systems that can handle the complexity of real business workflows. This focus on practical enterprise needs is precisely the kind of value proposition that sophisticated investors backing the latest AI funding round found compelling.

The first product, STARK RL envs, is a reinforcement learning environment specifically designed for AI agents. What makes STARK stand out is that it mirrors actual enterprise workflow scenarios — including common failure modes and edge cases — so that AI agents can train in conditions that closely resemble what they will encounter in live business settings. Before an AI agent is deployed in a high-stakes environment like financial processing, customer operations, or supply chain management, it can be stress-tested within STARK, learning from its mistakes in a safe simulation before being exposed to real consequences. This dramatically reduces the risk of costly errors in production and shortens the feedback loop between model training and real-world readiness.

The second product, Helix Evals, is an AI evaluation platform built to give enterprises continuous visibility into how their AI models are performing over time. Model drift — the gradual degradation of an AI model's accuracy and relevance as the world around it changes — is one of the most underappreciated challenges in enterprise AI deployment. Helix Evals addresses this by continuously monitoring model outputs, flagging inconsistencies, and providing actionable insights that allow engineering and product teams to course-correct before performance issues affect business outcomes. Given that many enterprises are now running multiple AI models across different departments and use cases, having a unified evaluation layer is increasingly non-negotiable.

The third product, EnterpriseOS Agents, is perhaps Deccan AI's most ambitious offering. This is an AI-native operating system layer designed to automate back-office and middle-office functions using intelligent AI agents. From processing invoices and managing vendor communications to streamlining HR workflows and automating compliance reporting, EnterpriseOS Agents brings the power of agentic AI to the administrative backbone of large organisations. Fortune 500 companies, which often struggle with the inefficiency of legacy back-office processes, represent a prime target market for this product. With enterprises increasingly looking to cut operational costs while maintaining accuracy and compliance, the demand for enterprise AI automation tools like EnterpriseOS is set to grow exponentially in the coming years.

Marquee Clients and a Growing Enterprise Footprint

When evaluating the credibility of any early-stage AI startup, the quality of its client roster is often more telling than any funding announcement. In this regard, Deccan AI has earned considerable respect. The company counts Google DeepMind and Snowflake among its current customers — two names that carry enormous weight in the global AI and technology industry. Google DeepMind is widely regarded as one of the world's foremost AI research organisations, responsible for breakthroughs ranging from AlphaGo to Gemini. Snowflake, on the other hand, is a leading cloud data platform company trusted by thousands of enterprises worldwide for mission-critical data operations. The fact that both of these organisations have chosen to work with Deccan AI is a powerful endorsement of the startup's technical credibility and delivery standards.

Beyond these headline clients, Deccan AI has onboarded approximately 10 enterprise customers and consistently runs several dozen active projects at any given time. While the absolute number of clients may appear modest, it is worth noting that these are not small pilots or proof-of-concept engagements. These are deep, ongoing partnerships with organisations that have rigorous vendor evaluation processes and high standards for performance. Each active project represents a complex, multi-dimensional engagement that involves data curation, model evaluation, and ongoing support — all of which demands significant resources and expertise. As AI funding news continues to spotlight infrastructure-layer companies, Deccan AI's client profile positions it as a serious contender in the enterprise AI services market.

Looking ahead, the company is casting a wide net for its next wave of growth. Beyond the North American market where it already has strong footholds, Deccan AI is eyeing expansion into Southeast Asia, Japan, and the Middle East — all regions where large enterprises are rapidly scaling their AI adoption and desperately need the kind of post-training and evaluation infrastructure that Deccan AI provides. The new Bengaluru office will serve as a strategic hub for this regional push, leveraging India's deep pool of AI talent, competitive operating costs, and growing ecosystem of tech-forward enterprises.

What the $25 Million Means for India's AI Startup Ecosystem

India's AI startup ecosystem has been on a remarkable trajectory over the past few years, and the Deccan AI funding round is yet another data point in an increasingly impressive story. The participation of A91 Partners — one of India's most respected growth equity firms — alongside global heavyweights like Prosus Ventures and Susquehanna International Group (SIG) reflects the kind of cross-border investor confidence that Indian AI startups have been working hard to build. The fact that this AI funding news is drawing international attention underscores how India is no longer just a talent reservoir for global AI companies but is increasingly home to AI-native startups capable of commanding Series A cheques from world-class institutional investors.

Prosus Ventures' continued involvement with Deccan AI is particularly noteworthy. The global technology investment group had already made an earlier undisclosed investment in the company, and its decision to participate again in the Series A signals deep conviction in Deccan AI's direction and leadership. Prosus manages one of the world's largest technology investment portfolios and has a well-established track record of identifying and backing high-growth technology companies across emerging markets. Its re-investment in Deccan AI is a strong signal that the company's performance and product trajectory have met or exceeded expectations since that initial cheque was written.

For the Indian startup ecosystem more broadly, Deccan AI's rise is also an important reminder that AI infrastructure — as opposed to AI application — is becoming a rich area for value creation. While much of the public conversation about Indian AI startups revolves around consumer-facing applications and SaaS tools, companies like Deccan AI are building foundational infrastructure that the entire AI economy depends on. This is a higher barrier-to-entry business with stronger competitive moats, and it is attracting the kind of serious, long-horizon capital that can fund the sustained R&D investment required to stay at the frontier. The AI World Organisation views this trend as central to India's emerging role as a global AI infrastructure hub.

Capital Deployment: Building Deep for the Long Game

With $25 million now in the bank, Deccan AI has outlined a clear and ambitious roadmap for how it intends to deploy the capital. The company plans to significantly scale its post-training data operations, investing in the systems, tools, and contributor networks needed to generate higher volumes of higher-quality training data across a broader set of domains. This includes a deliberate expansion into robotics data — a fast-growing segment as autonomous systems become more prevalent in manufacturing, logistics, and healthcare — as well as enterprise-specific data domains that are currently underserved by existing providers.

In parallel, Deccan AI will be making meaningful investments in research and development, particularly around the STARK and Helix product lines. The goal is to push the technical boundaries of what is possible in reinforcement learning environments and model evaluation, ensuring that the company's products remain ahead of an intensifying competitive landscape. On the enterprise side, the company will accelerate the buildout of EnterpriseOS Agents, working closely with its existing Fortune 500 clients to co-develop capabilities that address specific back-office automation challenges at scale.

The competitive landscape Deccan AI operates in is formidable. It competes with well-funded and well-known players including Scale AI — recently acquired by Meta — as well as Surge AI, Outlier, SuperAnnotate, and Mercor. However, Deccan AI differentiates itself through its exclusive focus on the post-training and production layers of the AI lifecycle, its deep integration with enterprise workflows through EnterpriseOS, and its remarkably diverse contributor network that spans over a million individuals globally. The company has reportedly achieved 10x business growth over the past year — a metric that speaks to both the strength of its offering and the scalability of its model.

As AI funding activity continues to accelerate globally, Deccan AI's $25 million Series A stands as a compelling example of how specialised, infrastructure-focused AI companies are capturing investor imagination and capital. The AI World Organisation, as the global apex body representing thousands of AI leaders, industry professionals, and enterprise innovators, will continue to monitor and report on AI funding news of this nature — recognising the profound role that well-funded AI startups like Deccan AI play in shaping the next generation of enterprise technology.