unbabel / backend-engineering-challenge Goto Github PK

In the README.md file there's an example of an input and an example of an output.
Are those two related?
In other words, for the provided input example, the application is supposed to write the provided output example?

The reasons why I ask this are:

• You mention a window size of 10 minutes but the provided output shows a window size of 14 minutes
• Unless I'm missing something about the challenge itself, the average_delivery_time field of the last 3 lines of the example output don't seem to make sense according to the provided input.

Thanks for your help.

Hi , am wondering about the input is it the static input that have been included or there is an external source like api for the input ?

Cheers,
Feras

{"timestamp": "2018-12-26 18:11:08.509654","translation_id": "5aa5b2f39f7254a75aa5","source_language": "en","target_language": "fr","client_name": "easyjet","event_name": "translation_delivered","nr_words": 30, "duration": 20}
{"timestamp": "2018-12-26 18:15:19.903159","translation_id": "5aa5b2f39f7254a75aa4","source_language": "en","target_language": "fr","client_name": "easyjet","event_name": "translation_delivered","nr_words": 30, "duration": 31}
{"timestamp": "2018-12-26 18:23:19.903159","translation_id": "5aa5b2f39f7254a75bb33","source_language": "en","target_language": "fr","client_name": "booking","event_name": "translation_delivered","nr_words": 100, "duration": 54}

Hi! I'm thinking in two approches that I would like to persue for this challenge, and I'll like to know if I can do both of them by creating two solutions, or any of the approches would be acceptable for you guys and I can create a single solution.

One: A CLI application that reads the events from a json file and generates an aggregated output file.
Two: A CLI application that listen to the events from a kafka topic and writes the aggregated output into a second topic.

Best regards!

I would like to get cleared about few unclear things in the challenge.

1. I think following line should be "we deal with".
   At Unbabel we deal we a lot of translation data.

2. Also the following "25,5" should be "25.5"

{"date": "2018-12-26 18:15:00", "average_delivery_time": 20}
{"date": "2018-12-26 18:16:00", "average_delivery_time": 25,5}
{"date": "2018-12-26 18:17:00", "average_delivery_time": 25,5}

3. "Feel free to, in your solution, include some your considerations while doing this challenge." <- is this mean to comment on my coding choices. Like why I do that, this way.. Is it?

Please let me know.

Hi there,
I'm working on the current challenge and i ran into an issue.
For the very first event, when calculating the moving average it returns 20 for the duration which is not the expected output. I saw the previous issue (which helped me to better understand the problem) but for this first one I'm still unable to wrap my head around it.

I appreciate any help :))

Hi guys!

I have a question regarding the window_size parameter.
In your example you inserted a window_size of 10 but in the output example the window_size is at least 13 minutes (18:11:00 to 18:24:00).
am I analysing in the wrong way?
Or the input and output example are not related?

Cheers.

I would like to clarify a couple of things:

1. What time unit is the duration in? The default mico-seconds or something else?
2. Do we get a test events.json file, or should we just use the example input file?

And finally, I think I may have found a typo. I think "we're instered in calculating," should be "we're interested in calculating."

1. What time unit of duration?.
2. What is the maximum/minimum window size?.

Before starting this I'ld like to confirm few things

1. Is the duration in seconds?
2. Past X minutes, is it with respect to current time (when the script is run), or the latest timestamp in the file?

I would like to know if we can use frameworks/platforms like spark streaming or other streaming frameworks.

I wanted to confirm whether input file is a static file with fixed number of events or a file with stream of events saving in periodic manner.

Thanks,
Madhu.

Hi!

Noticed that commit 9ea53b5 reverts changes done in past commits to an older version. To a state similar to commit 8a70cb2 where existed a typo in the name of the challenge and the expected output was different.

I wanted to ask if these changes were deliberate or intentional since the expected output changed and the input kept the same.

Thanks 🙏

1. If events.json is to be treated literally as JSON file then its formatting is bad since it has multiple root elements, which tempts me to consider its to be treated as text file with each line as a stringified JSON. Please confirm.
2. Unit of duration would add more clarity to data interpretation here. Perhaps, rename "duration" field to "duration_in_{unit}"? Please clarify what duration field unit is for now.
3. Regarding past X minutes- "from timestamp" information is not given. Could I consider from most recent events' timestamp from events.json file?

Hey guys, hope you are doing well.

The content of the file named events.json, suggests an incorrect format for a JSON file. It looks like an array, but we don't have the [], and also there is no separator between lines.

Should I understand that this is the correct structure and the '.json' the suffix is just less important information or the content suggested is really incorrect?