NYTech
People
Personfounder

Arvind Jain

Arvind Jain is an Indian-born technology executive best known as co-founder and chief executive of Glean, an enterprise search and AI assistant company, and as a co-founder of the cloud data security firm Rubrik [1][3][5]. Before turning entrepreneur he spent roughly a decade at Google, joining while it was still a pre-IPO startup and working as a distinguished engineer on Search, Maps and YouTube, following earlier engineering roles at Microsoft, Akamai and Riverbed Technology [6][7][8][9][13][14]. He holds a bachelor's degree in computer science from the Indian Institute of Technology, Delhi and pursued a master's degree in computer science at the University of Washington [10][11][13].

Jain has described his move into entrepreneurship as largely incidental: he left Google to co-found Rubrik in 2014 after a friend, Bipul Sinha, asked him to help start an enterprise software company, even though Jain's own background was in consumer products [13]. Rubrik grew rapidly, reaching more than a thousand employees within a few years, but Jain has said that around the time the company roughly tripled in size, productivity fell as employees struggled to locate internal information and identify the right colleagues to help them, despite the company running some 300 internal systems [14][18]. That experience, combined with his years as a Google search engineer, led him to conclude that no existing product could connect an organization's disparate SaaS systems and deliver a "Google-like" search experience inside a company, which prompted him to found Glean in early 2019 [14][18][4].

Jain has framed Glean's founding bet as a wager on transformer-based language models, made before generative AI entered wide public use; the company built what he describes as an early enterprise vector-search system, using models such as BERT pre-trained on individual customers' data to enable semantic rather than purely keyword-based search [14]. According to Jain, Glean's earliest and largest user groups were engineers and customer-support staff, with sales, legal and other knowledge-worker functions adopting the product later; he has cited a widely repeated figure that roughly a third of employee time is lost searching for information [18]. He describes the product's evolution from a document-finding search tool into a conversational "AI assistant" that synthesizes answers directly from company data, and more recently into "Glean Agents," which he positions as proactive, agentic systems rather than passive tools [13][16]. Jain identifies three technical requirements for enterprise AI adoption: strict adherence to existing user permissions, prevention of hallucination through citation-backed, source-grounded answers, and fine-tuning of models on company-specific data and terminology [16]. On broader industry trends, he has argued that closed models from providers such as OpenAI and Anthropic currently dominate enterprise use cases, but that open-source and distilled models tend to take over as applications reach scale, driven by cost, latency and, for regulated industries, data-control requirements [15].

Jain has also spoken about the transition from technical leadership to being a first-time chief executive at Glean, noting that Rubrik's cybersecurity product had a narrow set of internal buyers with whom his team built close relationships, whereas Glean is used by every employee at a customer organization, requiring the product to reach high quality before go-to-market efforts could scale [17]. He has said that becoming CEO forced him to learn skills outside engineering, particularly selling and working across marketing, HR and finance, and that he had to continually revise his leadership approach as Glean grew toward roughly a thousand employees [17][13].

Insights & ideas

Arvind Jain built Glean on a contrarian bet made in 2019, before generative AI was mainstream, wagering on transformer models and semantic search rather than keyword search, so the system could understand meaning and intent across scattered enterprise apps like Jira, Slack, and Google Docs [1]. He describes the shift from Glean's original role as a "librarian" pointing to documents toward an AI that synthesizes direct answers from company knowledge, arguing this requires solving three non-negotiable hurdles: respecting user permissions for security, preventing hallucinations, and selecting or fine-tuning the right models [1].

His core argument on trust is that hallucinations can only be solved by grounding every answer in verifiable company documents with citations, making the system transparent rather than a black box [1]. He also emphasizes fine-tuning models on company-specific jargon and data to create a personalized "AI teammate," and envisions AI becoming proactive and agentic, for example flagging outdated documentation automatically [1]. Ultimately, he frames enterprise AI's purpose as augmenting people as a force multiplier, not replacing jobs, freeing workers from tedious information-finding for strategic work [1].

Education

Media & appearances

In the news

This page shows public professional information only, each fact cited. Is this you? Corrections or removal within 24 hours, no questions asked.