Young Innovators @ Google – Manas Tungare
June 14th, 2011 | Published in Google Student Blog
Introduction: We recently launched Young Innovators @ Google, a new blog series highlighting the great work of Googlers who, not too long ago, were students like you. In their short careers, these engineers and product managers have had an impact on Google and our products. For today’s post, we sat down for a few questions with Manas Tungare, a software engineer at Google, who recently helped launched Voice Search for desktop.
Tell us about your academic background and your interest in HCI.
Tell us about your academic background and your interest in HCI.
Manas Tungare: Before I joined Google in 2009, I was at Virginia Tech getting my Ph.D. in human-computer interaction (HCI), and before that, a master’s in computer science (CS) from Georgia Tech.
I was drawn to HCI after encountering one badly-designed user interface after another. I figured I'd enjoy working in a field within CS where I'd get to design innovative user interfaces and implement them as well. As part of my Ph.D. program, I took psychology courses to understand how humans process information, and that knowledge has been tremendously useful for creating the kind of minimalist UIs for which Google is well-known.
Secondly, like Google, I strongly believe that software is a tool that empowers people. And, like Google, I believe that to do this, it must be free. Not only do we release a lot of our internal tools as open source libraries, but we also liberate our users' data so they are free to move to a competitor whenever they choose.
Once the company decided to launch Google Instant, I shifted away from my role as a prototyper to become part of the expanding engineering team responsible for the launch. I worked with several talented front-end engineers to implement features in a way that would scale to Google's billions of users (as opposed to my earlier prototypes that were not designed to scale).
When Google Instant launched in September of 2010, it was noted as the biggest change to the Google search results page in a decade. Being able to view results while typing means that users can refine their queries to get better search results, and now have to wait less to complete their tasks. It was incredible working on something so cutting-edge for my first project as a Noogler (a term for “new Googler”) with the support of other seasoned engineers.
Googlers are encouraged to come up with ideas outside their area of work, and build prototypes or demos to sell the idea to management. The 20% time that Googlers are allotted to work on non-primary projects is pretty well-known outside Google. In addition to that, some teams — like ours, Search Features — take a week off every year, and let engineers and designers regroup outside their regular work teams and try to solve real users’ problems with innovative user interfaces.
Every once in a while, our team holds mock-a-thons, where we are encouraged to present our designs for making some aspect of Search work better than it does now. We identify the major pain points that we have personally encountered and come up with solutions to fix them. Some are minor tweaks; others are major redesigns. After measuring the impact to users carefully, (mostly to ensure that our intuition is correct, and that we are not, in fact, making the user experience worse with a new feature), the feature goes live to all of our users.
There are several subtle angles to this comment: first, he said this without flinching, in a matter-of-fact way. Building two new datacenters for a new feature isn't something I would have expected someone to say that way. Secondly, being able to create a new feature that has such a high footprint is both exhilarating and humbling. Only at Google would this conversation take place with such enthusiasm and little hesitation!
Posted by Jessica Safir, University Programs Coordinator
I was drawn to HCI after encountering one badly-designed user interface after another. I figured I'd enjoy working in a field within CS where I'd get to design innovative user interfaces and implement them as well. As part of my Ph.D. program, I took psychology courses to understand how humans process information, and that knowledge has been tremendously useful for creating the kind of minimalist UIs for which Google is well-known.
Why Google? What makes working at Google unique?
MT: There are several reasons why I joined Google, the most distinguishing of which has nothing to do with technology. During my first internship at Google, when I was a Ph.D. student, I came to understand and appreciate the Google culture. It's a different way of thinking: engineers here think big. If a certain feature calls for a few hundred thousand machines, no one blinks. If a feature requires a hundred cars roaming the streets of the world taking pictures and capturing terabytes of data a month, no one blinks.Secondly, like Google, I strongly believe that software is a tool that empowers people. And, like Google, I believe that to do this, it must be free. Not only do we release a lot of our internal tools as open source libraries, but we also liberate our users' data so they are free to move to a competitor whenever they choose.
What was your first project at Google, and what impact did it have? How has your role evolved since then?
MT: Soon after I joined, I started work on what eventually became Google Instant. I was the sole engineer/prototyper on that project for several months, along with my manager, Othar Hansson, who came up with the original idea. I created several tens of prototypes; some of them were incremental changes over previous ones; others were completely different in appearance and functionality.Once the company decided to launch Google Instant, I shifted away from my role as a prototyper to become part of the expanding engineering team responsible for the launch. I worked with several talented front-end engineers to implement features in a way that would scale to Google's billions of users (as opposed to my earlier prototypes that were not designed to scale).
When Google Instant launched in September of 2010, it was noted as the biggest change to the Google search results page in a decade. Being able to view results while typing means that users can refine their queries to get better search results, and now have to wait less to complete their tasks. It was incredible working on something so cutting-edge for my first project as a Noogler (a term for “new Googler”) with the support of other seasoned engineers.
It’s great that you got to run with such a huge project right off the bat. Since then, in what other ways has Google encouraged you to be innovative?
MT: Google is very open to bottom-up innovation. My manager and I would approach random Googlers with the constantly-evolving prototype, and this let us gather early feedback which influenced the design. The final product went through several iterations based on all the internal responses, and each aspect of the user interface is carefully tuned to preserve the qualities that Google is known for — minimalism, speed, and simplicity.Googlers are encouraged to come up with ideas outside their area of work, and build prototypes or demos to sell the idea to management. The 20% time that Googlers are allotted to work on non-primary projects is pretty well-known outside Google. In addition to that, some teams — like ours, Search Features — take a week off every year, and let engineers and designers regroup outside their regular work teams and try to solve real users’ problems with innovative user interfaces.
Every once in a while, our team holds mock-a-thons, where we are encouraged to present our designs for making some aspect of Search work better than it does now. We identify the major pain points that we have personally encountered and come up with solutions to fix them. Some are minor tweaks; others are major redesigns. After measuring the impact to users carefully, (mostly to ensure that our intuition is correct, and that we are not, in fact, making the user experience worse with a new feature), the feature goes live to all of our users.
Any fun Google stories you’d like to share?
MT: The first time we showed an early Google Instant prototype to senior Search engineers, one of them commented that if we decided to launch it publicly, we would have to build at least two more datacenters to handle the additional traffic. He said this so casually, like one of us saying we’d have to buy a new computer because the old one was a little slow. An entire new datacenter just so you could launch a new feature? Wow.There are several subtle angles to this comment: first, he said this without flinching, in a matter-of-fact way. Building two new datacenters for a new feature isn't something I would have expected someone to say that way. Secondly, being able to create a new feature that has such a high footprint is both exhilarating and humbling. Only at Google would this conversation take place with such enthusiasm and little hesitation!
What are you up to now, and what do you see yourself doing at Google next?
MT: I'm having fun building new things that can be launched to millions of users! After Google Instant, I worked closely with Google researchers specializing in voice recognition & speech technologies. They were keen to add speech recognition capabilities to Google Search. When they approached the Web Search team, I agreed to work with them to launch this feature. And after just a couple of months, we launched Voice Search on Chrome for your desktop, as announced during this morning's media event with the other new search features. Now you can talk to your computer to search in addition to typing out your query.Posted by Jessica Safir, University Programs Coordinator