为影响而连接：印度人工智能领域的女性

A Report by Kalaari Capital From the moment we imagined artificial intelligence, we were captivated by its potential. Long before it became a buzzword, AIlived in the pages of science fiction and the circuits of futuristic droids. But today, it is no longer fiction. AI rewrites code,diagnoses illness, powers financial systems—and soon, will redefine entire careers and industries. India is not a bystander to this shift. We are the digital backbone for the world’s enterprises.At the same time, homegrown AIstartups have collectively raised over $5B+ since 2020—demonstrating that the next wave of deep-tech value creation may wellemerge from Indian corridors. This is the compounding power of progress. Much like how the discovery of DNA has transformed every field—from medicine toforensics to anthropology—AI, too, is rapidly infusing itself into the fabric of daily life. As thepace accelerates, the question is nolonger if AI will change the world, but whoseworld it will reflect. That brings us to a critical truth: women, who comprise nearly half the population, remain underrepresented in the very systemsshaping our collective future. Today, they account for barely one-fifth of India’s technology workforce. Their presence thinsfurther as roles become more specialized. This is not just a diversity issue—it is an engineering one. AI systems are only as good as the data and teams behind them. Whenthose perspectives are narrow, so are the outcomes. Algorithms trained on incomplete realities can undervalue female creditapplicants, overlook symptoms in women of color, and filter out résumés that don’t ‘fit’ the pattern. Inclusion is not a matter ofoptics—it is foundational to building trustworthy and equitable AI. Foreword From our conversations with practitioners, the barriers holding women back from AI/MLleadership in India include lack of early exposure to AI tools and curricula, limitedmentorship or female role models, and workplace cultures that reinforce existing gendergaps. Amid the excitement over what AI can already do, it is easy to overlook what it ismissing. That’s why we’ve put together Wired for Impact: Women in Ind(AI)—to spotlight thegaps, but more importantly, to identify what we can do about them. This report sets out to do three things: Show the data –Mapping where representation stands today, and thedistance to travel.Explain the stakes –Why diversity is not charity, but a design andperformance imperative.Spotlight the path –Through champions, case studies, and actionablerecommendations.123 My invitation is simple: treat inclusion like any mission-critical KPI. Audit it. Resource it.Iterate on it. Whether you are building a foundation model, enterprise workflow, or nationalskilling programme—ask yourself: Whose realities are encoded in the systems you ship? If we get this right, India will not only secure its seat at the forefront of AI innovation—wewill ensure that the intelligence we unleash is truly representative of the country and thehumanity it aims to serve. VANI KOLAManaging Director,Kalaari Capital Before The FirstPrompt Methodology - Roles (1/2) This report began with a simple butimportant question: Where are the women in AI?Not in broad statistics or policy speeches—but in real teams, solving real problems, across India’s fast-growing AI/ML ecosystem. To explore this, we focused on women professionals working in roles that directly contribute to the AI value chain—across datascience, engineering, model development, and applied research. Using LinkedIn Sales Navigator as our primary discovery tool,we identified a curated sample ofprofessionals across 36 AI/ML-relevant job titles, grouped into five broad role types: Machine LearningRoles (e.g., ML Engineer,ML Researcher, MLOpsHead) Data Engineering &Architecture (e.g., DataEngineer, Senior DataArchitect) Data Science Roles(e.g., Data Scientist,Chief Data Officer, NLPEngineer) Deep Learning Roles(e.g., Deep LearningEngineer, CV Engineer,Robotics Engineer) Artificial IntelligenceSpecialist Roles (e.g., AIEngineer, PromptEngineer, GenAI Engineer) For simplicity, we grouped AI, ML, and Deep Learning under one functional umbrella. While each has its own academiclineage, they increasingly converge in industry—often sharing talent pools, project scopes, and underlying infrastructure. Meanwhile, Data Science and Engineering were treated as a single category, given their interdependence in the AIdevelopment lifecycle. From data readiness to experimentation and feature engineering, this combined track representsthe essential input pipeline without which intelligent systems cannot operate. Methodology - Organisations (2/2) We then analysed representation across five distinct organisational archetypes that shape India’s AI workforce: This mapping provided a window into patterns of representation across functions, career stages, and sectors. While the dataset is not exhaustive, itreflects directional insights and recurring theme