Technical Interview Questions

Kai W. - "The user awakes Alexa by saying the "Alexa word" Device starts recording audio from the user, and streaming it to the Alexa Cloud Service The recorded audio stream is sent to the ASR (Automatic Speech Recognition) system ASR returns the words and the order of the words in the audio stream, and sends this information to NLU (Natural Language Understanding) NLU returns the intent of the audio stream, and sends it to Alexa Skill If the intent triggers a "play" directive (e.g. turn"See full answer

Glados - "First, why are we building the system? What are the requirements we must meet - Do customers care about data being read/written from a single set of tables or multiple tables, do they care about cost allocation, do they need backups/storage to happen slow or fast, do they foresee a lot of data movement from one table to another? Based on this, there are a few options using some common design choices. Shared disk, or Shared nothing. Both have pros and cons. Based on latency, performance, and cos"See full answer

James H. - "I broke the vehicles down by regions in the world. Designed a wireless api interface to sync with a database backend that linked to google maps through another service and api."See full answer

Anirudh G. - "Questions for interviewer: How many team members are there? How much is the bandwidth required for each user? As of now radio-frequency (RF) signals are used to communicate with Moon. However in order to create a high speed network, we should be using laser communications. NASA has done research on it and results have proved that it is 10-100 times better than RF signals (around 100 Gbits/second). Geosynchronous satellites must be used to send communications and data back and forth betwe"See full answer

Mehul K. - "Clarifying questions I asked? Are we trying to track any specific type of metric ? Like a Business Metric, Engagement Metric ? No. In general as a admin of a city at Uber what would you track ? Okay. (After collecting thoughts) Revenue from rides per day No of rides per day in the city % of drivers completing more than 10 rides/day, 5 rides/day, 1+ ride/day."See full answer

Bini T. - "I understand this is more focused on ML. However, I have a system question. If users allow us to access their location, or they send location via text box, could we use CDNs for the search without hitting our database? We only query the database when we have zero information on location. Other questions: does embedding always guarantee information on location? Do we discharge the user images after we return a prediction? I heard the feedback that we should keep it for future learning. What would"See full answer

Ven E. - "I saw this question in one of my interviews."See full answer

Rjj - "I tried solving this question and here is the recorded video for the entire solution - https://youtu.be/G_LIbTp58XA Feel free to comment here or on the video for further discussion."See full answer

Rashmi T. - "It depends on how computers are distributed. Are they located within same network (behind same switch or router) or distributed across globally? Also, it depends upon if there are any security related encryption or decryption algorithms being executed. Similarly, there are so many factors involved in it. However, if a system is highly responsive then this time might be less than 1 second. For example, any google search responds in less than 1 sec. It is an interesting question and I am eager t"See full answer

Y C. - "One way to develop project metrics is around what problem the project is trying to solve within a particular timeframe, at a cost of x dollars and with y quality Scope: Did the project deliver as per the problem statement - over and beyond Resources metrics: Planned resources were utilized to contribute to the success Budget: Did the cost of the project implementation be under the budget or over? Schedule: In-flight milestones were completed on time or not as per the plan Was the proje"See full answer

Laura S. - "Clarifying questions: is this a brand new product, or are we improving an existing one? (i.e. are we going to have to migrate an existing codebase or are we starting from scratch?). are we resource-strapped (e.g. # engineers, time)? are there any specific priorities for the product, or should i leave it open-ended? Assume: brand new product, well-resourced, no specific priorities. It seems that there are two sides to this question: 1) technical evaluation of different languages, and 2)"See full answer

Catherine Z. - "Clarifying Questions Have we confirmed this was a server side issue and not client side or network related? Interviewer confirms this was a server side issue Do we have additional information on whether this was region specific, user specific? Interviewer says you can make some assumptions Introduction I am going to focus on a few potential factors and describe how they could lead to system downtime: Data Center Power Outage Software Deployment Software Bug P"See full answer

Anjaly J. - "Goal of the engine : 1. Recommend the "right" videos 2. Make users watch the videos continuously The design consists of 4 different components : Video database (corpus): Consists of million of videos. Recommendation Engine : Based on User history (watch and search history, when enabled) coupled with user context (country, time of the day), the millions of videos are filtered to hundreds of videos and passed through the recommendation engine Ranking : Apart from user history and contex"See full answer

Sourabh G. - "Functional Requirement Monitor health, metrics Alert in case of failure/anomaly Visualize the live health Analyse machines on periodic basis Non Functional Should not exert load on machines low latency Highly scalable Logs/Metrics Gathering push - machine gather and send to system and low priority background thread along with batching pull - heart beat check (for offline machines) Processing Real time streaming using Kafka/kinesis + Flink TimeSeries database for stor"See full answer

Rolemberg J. - "I faced a problem about components re-rendering's and unnecessary requests API's which was causing performance complications in my applicactions. I had a structure that create, edit and delete task, so GET, POST, PUT and DELETE API's methods request was necessary and bring that to compliance without compromissing the performance is hard. I started involving componentes and async functions into the useMemo's and useEffect's to have more control, another improvement was take be careful with global"See full answer

Wizzy B. - "URL> DNS over UDP> IP address > TCP handshake > HTTP(SSL handshake) req, responses, websites architectures for any query DB, servers."See full answer

Astro S. - "what is NUC? "See full answer