Google Technical Interview Questions

Aditya S. - "From Product Management perspective: I would first start with some clarifying questions. Is this algorithm to be used across Play Store for all kinds of apps or for some specific catgeory of apps What is the major objective. Are you entering a new market and hence objective is to capture market OR already in an existing market and want to maximise profit (since Google, most likely second objective) I would like to experiment with algorithms and see results before finalising. This will nee"See full answer

anush - "@Jacob, your videos are really good and helpful. I want to know about the tool, you use for explaining the design in the videos."See full answer

Jasmin R. - "Let's imagine you want to open Google Maps, you type maps.google.com and hit Enter. Here is what happens: You type maps.google.com into the address bar of your browser. Browser completes the address, for example, maps.google.com → https://www.maps.google.com [Address decoding] → browser breaks down the URL into parts ‘https://’: protocol → how the browser should connect to the website Default is HTTP (Hypertext Transfer Protocol) which shows up as ‘ http:/"See full answer

Adarsh S. - "Thanks for the question! This is one of my favourite topics because it’s at the core of what we do as PMs DAILY: making the complex simple to drive alignment. I’ve got a structured way to approach this, and I’ll use it to explain payment aggregators - like Plaid - and how they enable Open Banking, with a twist about a technical hiccup I’ve seen in practice. Let’s dive in. I start by assessing the audience. Say I’m talking to a marketing manager at Transak - someone sharp but not techni"See full answer

Mohit G. - "5 reasons Gmail search is real time Gmail search needs to be 100% accurate Does not have significant caching / hot searches Less hardware used as user info cant be sprayed across machines Gmail storage is different, as needs decryption on every"See full answer

Srini K. - "performance issues and sudden spikes on input requests by scaling techniques and optimization."See full answer

DM - "Clarifying my assumptions first, i.e. "Technical contributions are not just writing or reviewing code". Answer : My contributions to any program I lead as a TPM are as follows : Inputs towards design, architecture review, mapping requirements to the proposed design, reviewing the implementation strategy w.r.t scalable solution, Writing core user-centric test scenarios, Validating proposed design vs implementation estimates vs initial planning. etc... Example : Situation: We are a"See full answer

S G. - "If it's a product end to end, then there's a lot to say. If it's just a feature on top of a product then I will focus on the complexity and how I break complex tasks into small milestones to partner with XFN teams to nail it. Focus on situation, task, actions and results"See full answer

Surbhi G. - "Deep Learning is a part of Artificial Intelligence, it's like teaching the machine to think and make decisions on its own. It's like how we teach a child the concept of an apple - it's round, red, has a stem on top. We show them multiple pictures of apples and then they understand and can recognize an apple in future. Similarly, we feed lots of data to the machine, and slowly, it starts learning from that data, and can then make relevant predictions or decisions based on what it has learnt. A co"See full answer

Hamilton D. - "no"See full answer

Aaron W. - "Clarification questions What is the purpose of connecting the DB? Do we expect high-volumes of traffic to hit the DB Do we have scalability or reliability concerns? Format Code -> DB Code -> Cache -> DB API -> Cache -> DB - APIs are built for a purpose and have a specified protocol (GET, POST, DELETE) to speak to the DB. APIs can also use a contract to retrieve information from a DB much faster than code. Load balanced APIs -> Cache -> DB **Aut"See full answer

Sim S. - "Started with user-groups to identify the differences world wide: speed of internet, how high end devices are The interviewer just wanted to dive into the questions of latency & the technical fast. I then said let's separate out server vs client. On server side, cannot have australian users hitting california servers for data. So must be some sort of global availability of caching. And ensure rapid read out of that cache so something like a redis solution would help. Then switch gears to cli"See full answer

Karthik M. - "Assumptions & Clarifications Users have access to all Google Workspace applications like Google Drive, Gmail and the same photo editing tool - let us call this tool - Google Photo Editor (GPhoto). Users can share the images like Google Docs - giving people read-only, read-write / edit access Goals Google's vision is to organize the world's information and make it accessible. GPhoto's mission / goal is to organize the world’s photo and make it accessible to others to view, edit or"See full answer

Kanika - "https://www.freecodecamp.org/news/what-happens-when-you-hit-url-in-your-browser/"See full answer

Praful B. - "Define the goal of the experiment. What do you want to test? Are you trying to improve conversion rates, increase engagement, or something else? Identify the variables to test. What are the different options that you want to test? For example, if you are testing different call-to-action buttons, your variables would be the different button text and colors. Set up the experiment. This involves creating two versions of the app or website, one for each variation of the variable that you are testi"See full answer

Anonymous Muskox - "We will use online prediction instead of batch prediction because we can't compute beforehand. We have to do real-time image processing. For model deployment, we will use the model in service because we would want minimum latency and model in service to make sure that there are local network call rather than remote procedural calls."See full answer

Marwan O. - "My company was acquired and I was responsible for planning and managing a transition & integration of custom real-time time-series data and event capture software. I worked with lab scientists (who were running the instruments from which the data would be captured) to understand their unique needs compared to the use cases the software was originally written and deployed for. I wrote proposals for integrating specific components of the software and abandoning others, receiving review and a"See full answer

AnonymousAxolotl - "Depends on what the ML model. If it's computationally heavy server side is best. If it's not, client side could be feasible, think of Apple Photos doing object recognition of photos on device, in this instance, the photos are on the device a long time and users are fine with waiting. So the use case important to consider. However broadly speaking, server side is most likely the best solution especially for this many users. Easier architecture to implement Example - client side you take"See full answer

Parth P. - "Establish more servers on regional bases or locally and move network traffic to these servers instead of keeping on one server."See full answer