Amazon Business Intelligence Mock Interview: Duplicate Products

Подписаться 46 тыс.

Просмотров 78 тыс.

50% 1

Today I'm joined by Shashank who is a data scientist and business intelligence engineer at Amazon!
Question page: www.interviewquery.com/questi...
Want to be featured in the next mock interview video? Apply here: airtable.com/shrdQrwKK7xxGLm6l
👉 Subscribe to my data science channel: bit.ly/2xYkyUM
🔥 Get 10% off your next data science interview prep: www.interviewquery.com/pricin...
❓ Check out our data science interview course: www.interviewquery.com/course...
🔑 Get professional coaching here: www.interviewquery.com/coachi...
🐦 Follow us on Twitter: / interview_query
Shashank and I work on a business case study problem that is asked pretty frequently for Amazon business intelligence engineers and analysts. It involves breaking down a vague problem in a database into a solvable solution that makes sense and can scale for Amazon's purposes. At the end we go over feedback and tips and tricks for the next one!
Quick Links:
0:35 - Shashank Intro
2:55 - Business Intelligence Case Question
11:55 - Feedback and Tips
More from Jay:
Follow me on Linkedin: / jay-feng-ab66b049
Read my personal blog: datastream.substack.com/
Find me on Twitter: / datasciencejay
Read more:
The Amazon Business Analyst Interview: www.interviewquery.com/blog-a...
Amazon Business Intelligence Engineer Interview: www.interviewquery.com/blog-t...
Amazon Machine Learning Questions: www.interviewquery.com/blog-a...

Наука

Опубликовано:

4 авг 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 55

@prasshanthvp1082 3 года назад

My 2 cents: Product similarity is a complex problem to solve even for established top 4 companies. Image matching for product requires data science solution but that's just one small part. String matching could be managed better with data science solution, but if all you have is SQL, you can still create a good baseline using string functions like regex fns for matching. Your product matching confidence increases when you introduce more parameter to match. Dimensions like weight, size, height as pointed by someone in comments are good examples, Reviews are not simply due to sheer noise and overhead involved in parsing long winded strings. Think of what parameters will you look at when you buy a smartphone, a good laptop, a nice pair of earphones and retain only those parameters which makes sense to match two products.

@dhaneshshirodkar365 3 года назад

This is more like a Data Science interview

@upskill9257 2 года назад

I've been looking for realistic data science/analytics case study discussions for some time. This has been by far the most relatable one for me. Please add more content like this.

@manoharlakshmana6171 4 года назад

This is what we need !!! Case study mock interview !!! Thank you so much !!! Hope you will upload more such videos.

@saumya4582 3 года назад

Thanks for this and many other videos Jay! I was looking for ways to get to know about the interview questions asked to work upon and stumbled upon your channel. Super helpful as I had been rejected at my dream company's interview and was starting again to work on my gaps to be able to apply again for the interviews.

@mallikatiwari252 4 года назад

Super Helpful ! Can you please do similar video on Amazon Business Analyst?

@manoharlakshmana6171 4 года назад

Please Have a video on : Suppose an ecommerce website is planning for a flash sale. How would you select which products to put on the sale.you can add more questions to it and then followed by A/B testing what kind of statistical tests would be used and where.

@jaganshemthegay5490 4 года назад

De duplication is great master data quality exercise that needs domain experts inputs and study of the features and attributes.. Once de duplicated transactions can be tied down to unique product is which gives better analytics

@xiaoyangwu9270 4 года назад

Thanks for all the advices on implementing ML/DS algorithms into data engineer/Data warehouse part of the task, I am always wondering how those two can be combined when comes to solving problems togather

@jaganshemthegay5490 4 года назад

Consider sone master data attributes and combination of these attribute values gives the uniqueness to derive .. color , size , iOS version , first available date etc and some other features and do deduplication and map all product names to one unique artificial key

@EmWatsonable 4 года назад

Thanks for sharing! This is super helpful :)

@nitishnayak6834 4 года назад

Can we use dimensions and weight of the mobile. Definitely similar products will have similar description and design

@SoraAICreator 4 года назад

This is really helpful. Thanks a lot. Can you do a mock Product Analyst Interview as well?

@bibiworm 3 года назад

This is a very classic interview question for e commerce business. 👍 But this mock interview only scratched the surface of it. I have been asked the same question. I was grilled on similarity measures, cnn for image processing, rnn for nlp, etc.

@upskill9257 2 года назад

Can you share some additional parameters that were not covered here?

@yashsolanki069 4 года назад

You are doing really great sir👍

@porwalabhishek5 4 года назад

This is so helpful

@ambarishdeshmukh143 3 года назад

Thank you for the Video and the contents. Really appreciated.

@alexz7032 3 года назад

Awesome stuff! thanks

@kalpeshsajnani9952 3 года назад

Thanks a lot for such a informative video

@tigerbear3038 2 года назад

SKUs often differ from merchant to merchant. You have to look at the model number or UPC

@chitranjantrivedi6008 3 года назад

Bro - Do we need to know Data Structures and Algorithms to crack Business Intelligence Engineer interviews at Amazon?

@pushkarajpalnitkar1695 4 года назад

Hi, Jay and Shashank, thanks for the video. I have one doubt though. Did he have to give data structures and algorithms' interview for the job?

@iqjayfeng 4 года назад

Nope I don’t believe so

@tonestrada2000 Год назад

Do you have sample data for this? This seems a good case study for beginners like me

@karthikjosyula8002 4 года назад

wouldn't equivalent classes solve the problem of de-duping the classes Apple iPhone 10 = iPhone X... We can store these classes in under same ID and use the ID. Even tracking seller details where same seller has same product, labelled differently, posted multiple times.

@mr.garammasala1398 3 года назад

Fuzzy look up/ fuzzy grouping works here.

@kushagrasen1848 3 года назад

we can also take the product dimensions

@asthasrivastava9564 3 года назад

Quick question - Has anyone faced a question like this for the BIE role (may or may not be Amazon)?

@vincentpun123 3 года назад

nice

@fionaw9942 4 года назад

Does BIE need to have ML knowledge?

@iqjayfeng 4 года назад

Depends on the role but usually no

@mitchieyy 2 года назад

if the product have the same unique id such as SKU UPC they definitely get under the same listing when you list them. as an amazon seller i know that sellers get the product from the same manufacturer and they use their own brand name. that’s why you see the same product over and over again on amazon but they will have their distinct brand name, like kirkland and huggies get their diapers from the same manufacturer. with the big companies they manufacture the product slightly differently and package them differently. but with small amazon seller they don’t do all that and their product photo looks the same. In this case i don’t think they could be identified as the same because they are different brand. So to identify true duplicate that have similar image most importantly you need to check their brand are the same and then their key product description such as number of unit weight color size version are the same should they be considered possible duplicates.

@julialjo7921 3 года назад

“Nobody knew how to solve these tasks” and then Rajesh came all the way from India to save them.

@FPrimeHD1618 2 года назад

This helps out a lot, much appreciated. I have an interview for a Sr BIE position next week. First time interviewing with Amazon and I'm freaking out lol.

@extcresources531 2 года назад

What was the result of your interviews? What all were asked? If you could share it will help a soul. Thanks mate

@FPrimeHD1618 2 года назад

@@extcresources531 I actually was offered a really good role as a data scientist in my current company that I couldn't pass up, so I stopped all interviews.

@saniyajaswani4462 2 года назад

His solution is not scalable If we make such algorithm, how do we scale for 10000 products. Amazon does least manual intervention on duplication product detection

@fredbradshaw4308 2 года назад

I was thinking simple Regex 😂

@allison-hd1fg 2 года назад

I think this case study is more for data scientist role. you sure this is a case for BIE?

@iqjayfeng 2 года назад

Definitely for BIE roles at Amazon specifically

@toekneema 3 года назад

what is a DFIDF? 7:40

@nanfengbb 3 года назад

Should be TFIDF, a technique commonly used for NLP.

@nhrakib1572 3 года назад

hi dear how are you

@annachan6706 3 года назад

why not where product name LIKE '%iphone%'

@extcresources531 2 года назад

Cos that would give you same brand but phones of different models.. iphone 6 iphone 7 iphone 8 etc..

@iqjayfeng 3 года назад

Thanks for watching! If you want to prepare for Business Intelligence roles, then Interview Query has you covered! We have diverse questions to prepare for your Business Intelligence interview and stay ahead of the curve. www.interviewquery.com/p/business-intelligence-interview-questions