Mid-Stride at the Center of Everything
Somewhere between your morning alarm and your first meeting, Prachi Gupta's work is already running. The email that woke you. The calendar invite you just accepted. The chat message you're composing. She is VP of Engineering at Google for Gmail, Google Chat, Google Calendar, and Google Tasks - four products so embedded in professional life that their outages make headlines and their new features generate corporate memos.
That is the scope of the job. But the nature of it is shifting. Gupta is not simply maintaining infrastructure at scale - she is overseeing the translation of those products into something different: AI-native tools where Gemini doesn't just assist, it anticipates. Where Gmail doesn't wait to be searched but surfaces what you need before you ask. Where Calendar doesn't just hold your schedule but negotiates it.
This is the mandate she stepped into when she moved from YouTube to Google Workspace in 2023. And it's a mandate she has been prosecuting at pace ever since.
From IIT Roorkee to Inbox Zero
Gupta's path into Google's inner engineering orbit follows the kind of arc that looks inevitable in retrospect and was anything but in the making. She earned her undergraduate degree in Computer Science and Engineering from the Indian Institute of Technology, Roorkee - one of India's oldest technical universities and a pipeline for some of the industry's most consequential engineers. She followed that with a Master of Science in Computer Science from The Ohio State University, bridging two continents and two distinct engineering cultures.
Her career in the industry ran through Intel and Yahoo before she arrived at Google. Then came the eight-year chapter at YouTube.
"The value of an AI assistant is measured by its ability to understand context and complete complex tasks on your behalf."- Prachi Gupta, Google VP Engineering
From February 2015 to April 2023, Gupta served as VP of Engineering at YouTube - a platform that streams a billion hours of video daily and operates at a technical scale that makes most engineering problems look quaint by comparison. The infrastructure challenges alone would occupy most engineering careers. The product evolution - from desktop to mobile to smart TV to shorts to live to shopping - demanded something more than competence. It demanded the kind of institutional fluency that only time and sustained attention can build.
Eight years is long enough to see multiple product cycles through. Long enough to build and rebuild teams. Long enough to have an opinion on what actually works and what is just impressive-sounding on a slide deck.
In April 2023, she made the pivot to Google Workspace - first as VP/GM for Google Core User Protection, then expanding her scope to lead engineering across all of Google Workspace's communication and time management products. The move was a shift in domain, but not in scale. If anything, the audience got larger.
Four Products. One Mandate.
The four products Gupta's engineering organization is responsible for are not new. Gmail turned twenty in 2024. Google Calendar has been the default office scheduler for the better part of two decades. But their age is precisely what makes her current work interesting - and difficult.
Rearchitecting tools that people use on muscle memory is a different engineering challenge than building something new. Users notice every change. Every omission. Every millisecond of latency that wasn't there last week. The bar for disruption is higher, not lower, when the product is already ubiquitous.
Gmail Enters the Gemini Era
The phrase "AI-first" gets used until it means nothing. Gupta's version of it is more specific. At Google Cloud Next 2026, she framed the shift in terms of outcomes rather than capabilities: momentum had moved from demonstrating what AI could do in the abstract to measuring what it actually accomplishes for people sitting in front of their inboxes.
The Workspace Intelligence initiative she championed is built on a particular premise - that a context-aware assistant is categorically different from a reactive one. Most AI tools respond to prompts. Gupta's vision for Gmail is something that reads your calendar, knows your recent emails, understands your working patterns, and acts before you ask.
Gmail's AI Transformation - Gupta's Framework
The Gemini Enterprise Agent Platform and the Agent Designer - both of which Gupta highlighted at Google Cloud Next 2026 - are the infrastructure layer under this vision. They allow Workspace to function not just as a suite of tools but as a coordinated agentic environment where AI can move across applications without losing context.
This is not a small engineering project. It involves rewriting how identity, permissions, data access, and real-time signals move across a platform used by more than three billion people. Getting it right matters at a scale that is genuinely hard to hold in your head.
Building Teams When Everything is Remote
In 2020, when every engineering organization was improvising distributed leadership, Gupta did something practical: she went public with her thinking. At the Executive Leadership Conference (ELC), she co-presented with Minal Mehta on "Building & Managing a Diverse Team in the Age of Social Distance" - a session that landed because it was specific rather than aspirational.
The session's underlying argument was that distributed work doesn't just change logistics - it changes who gets visibility, who gets mentorship, who gets promoted. The engineers who had learned to be present in person now had to learn to be present at a distance, and the leaders who had built their intuition around proximity needed to rebuild it around signals they could actually read through a screen.
For Gupta, diversity and inclusion are engineering problems as much as cultural ones. A team that looks like only one thing produces software that works for only one kind of user. Given that her products serve billions of people across dozens of languages and hundreds of cultural contexts, the stakes of that argument are unusually concrete.
"Momentum has shifted to focus on concrete outcomes rather than abstract potential - specific instrumentation driving the agentic transition focused on outcomes."- Prachi Gupta, Google Cloud Next 2026
Managing engineering at this scale requires holding two things at once: the long horizon of platform architecture decisions that take years to play out, and the week-to-week velocity of product teams shipping to users who will notice every change. Gupta has been doing that for a decade across two of Google's most consequential product families.
The Work That Speaks
- Led engineering at YouTube - one of the world's highest-traffic video platforms - for over eight years, spanning the shift to mobile, the rise of Shorts, and the integration of machine learning into recommendations
- Drove the Workspace Intelligence initiative, giving Gmail, Docs, Drive, and Chat a shared real-time AI understanding layer for the first time
- Oversees engineering for Gmail - the world's largest email service at over three billion users - through its most significant architectural shift in a decade
- Presented the Gemini Enterprise Agent Platform and Agent Designer at Google Cloud Next 2026, marking a new phase in Workspace's AI roadmap
- Advocated publicly for distributed team management and diversity in engineering leadership through ELC, reaching a broad audience of technical leaders
- IIT Roorkee graduate - one of India's most selective technical institutions - and MS in Computer Science from The Ohio State University
What She's Actually Building
The most interesting version of the question "what does Prachi Gupta do?" isn't organizational. It's architectural. She is building the scaffolding for how AI agents will exist inside the tools that knowledge workers already trust - not as add-ons, but as native participants in the flow of work.
That is a harder problem than building AI from scratch. Users have expectations about email. They have muscle memory about calendar invites. They have anxieties about who can read their messages and what happens to that data. The engineering has to be right, and the trust has to be earned, and both of those things take time at a pace that startup analogies don't map onto.
Gupta has spent her career at companies that operate at scales where individual engineering decisions become policy for hundreds of millions of people. The choices she and her teams make about how Gemini reads context in Gmail, about what the AI can and cannot do with Calendar data, about how Tasks integrates with the agentic layer - those choices will shape how a significant portion of the world works, communicates, and organizes their time.
That is the job. She is mid-stride in it.