Amazon Interview Question
SDE-2sCountry: India
Interview Type: Phone Interview
Search for all the data rows in table corresponding to trials made by the customer in an attempt to use Alexa with Phillips light bulb.
This could mean the rows for example, where the Alexa commanded the bulb to be switched on, but in reality the bulb didn't switch on due to issue and in database, there was a row made for this, when next time the Alexa commanded for it to be switched off then the program could not really find a row where the data matched with the ON state and hence an exception could have been generated.
The fix program could identify such data rows in tables where actual transactions didn't happen in reality and delete those rows.
The data scrubbing solution will comprise of following parts:-
1. A support service platform (API) which will broker the call between customer and AMZN support and then between customer and 3rd Party company. The API will provide entry points to capture all valid inputs as desired to captured for the support call.
2. The support services API will interact with data scrubbing API (Google Refine etc.) which will help eliminate, clean undesired data by using regular expressions and other algorithms.
3. Support services API will be responsible to persist the support related information in the data store
4. An API gateway (AWS, Azure, GCP) will be used manage API requests at AMZN scale which will also be responsible for Authentication as well.
5. A master agreement needs to exist between 3rd party and AMZN for leveraging AMZN support platform for service calls or alternately Google Voice API's can be leveraged to analyze support calls and perform actions as needed.
The data scrubbing solution will comprise of following parts:-
1. A support service platform (API) which will broker the call between customer and AMZN support and then between customer and 3rd Party company. The API will provide entry points to capture all valid inputs as desired to captured for the support call.
2. The support services API will interact with data scrubbing API (Google Refine etc.) which will help eliminate, clean undesired data by using regular expressions and other algorithms.
3. Support services API will be responsible to persist the support related information in the data store
4. An API gateway (AWS, Azure, GCP) will be used manage API requests at AMZN scale which will also be responsible for Authentication as well.
5. A master agreement needs to exist between 3rd party and AMZN for leveraging AMZN support platform for service calls or alternately Google Voice API's can be leveraged to analyze support calls and perform actions as needed.
The data scrubbing solution will comprise of following parts:-
1. A support service platform (API) which will broker the call between customer and AMZN support and then between customer and 3rd Party company. The API will provide entry points to capture all valid inputs as desired to captured for the support call.
2. The support services API will interact with data scrubbing API (Google Refine etc.) which will help eliminate, clean undesired data by using regular expressions and other algorithms.
3. Support services API will be responsible to persist the support related information in the data store
4. An API gateway (AWS, Azure, GCP) will be used manage API requests at AMZN scale which will also be responsible for Authentication as well.
5. A master agreement needs to exist between 3rd party and AMZN for leveraging AMZN support platform for service calls or alternately Google Voice API's can be leveraged to analyze support calls and perform actions as needed.
Looks like Product question,
- Vib February 02, 20191. Educate customer about not sharing the private data.
2. Have agreement with Philips asking for an API to query the voice samples anytime and check from Amazon side. Since its support information, they should not have any objecting to sharing.
3. Work with Philips to understand the issue why customers had to call in first place, if there is a bug force Philips to release fix asap.