About Myself Flashcards
Tell me about yourself
I have been working in analytics and data science for about 4 years, and I primarily focused on optimizing business processes and improving customer outcomes.
My experience ranges from working on customer segmentation and behavior analysis to developing models that enhance business performance, for example reducing customer churn or improving sales operation.
Most recently, at Tesla, I focused on using data science to boost sales performance, optimize costs, and get deeper insights into customer behavior, and product perception, especially the AI-based products.
I think one of my key strengths is being able to bridge the gap between technical and non-technical teams. I enjoy working closely with non-technical stakeholders to understand how they operate, what their priorities and needs are, identify the opportunities, and then translating those into actionable data challenges. In the past this approach helped me deliver a lot of value, and I am excited to bring this skill into this new role.
In my data science free time I am a big fan of fiction an fantasy - I have watched the GOT 3 times and read all the books! I also like spending time outdoors with my dog, hes my best buddy and it really helps me clear my head after a long day. And lately you can find me a lot in places like IKEA, Praxis, and Karwei, because my husband and I recently bought a house in Ouderkerk, so DIY seems to be my new hobby these days!
That’s a bit about me!
I am really excited to get to know you both and hear more about the team and what you’re working on!
Why did you leave Tesla
I have been at Tesla for 7 years and I grew a lot and I learned a ton during my time there.
But lately, for a while, I felt like my growth was starting to stagnate and then Tesla announced the layoffs and I realized that the priorities will shift even more because I already lived through 2 layoff cycles at Tesla,
so I know that it means going back to the basics, focusing on the short team goals and quick wins, one day at a time, which for me me and my stakeholders meant de-prioritizing projects that are focused on building and improving and completely focusing on “running” the operations.
And i realized that it’s not aligned with where I want to go because i want to learn and grow and innovate
So i decided that it’s best to part ways
What is your main weakness?
I need a lot of stimulation and I find it difficult to stay motivated when for a prolonged periods of time I feel like im not growing
Why would you like to work at this company and this role?
I think the focus of the company aligns with my interests. I am passionate about customer analytics and data-driven innovation. I like that the company sees a lot of value in data analytics and data science and i like that there is a big data scientist hub.
I also like that the company’s mission and the culture feels very human-centered, with a strong focus on the customers and employees, and of course society and environment.
From what i have seen so far it seems like a very nice place to work, with a lot of focus on employee development, which for me is very important too, as im still very early in my career.
Why do you like this role?
I am passionate about customer analytics and AI.
And even though when i was applying I dindt know yet that this role would be specifically in the banking, i do find credit risk modelling very interesting as well
It also sounds like there is a big data scientist hub at the company
Can you tell me about your experience with MLOps?
Although i dont have direct experience with it because we didnt really have this infrastructure, I understand the functionality and I am eager to learn on the job.
I understand that MLOps is all about automating and streamlearning the machine learning lifecycle and making sure that the models run smoothly and can be updated easily, t’s like DevOps but focused on ML models
I am also happy to take some course prior to starting, if that’s preferred
MLOps is all about automating and streamlining the machine learning lifecycle, from deploying models to monitoring them in production. It’s similar to DevOps but focused on ML models, helping ensure they run smoothly and can be updated easily.
Can you tell me about your experience with AWS SageMaker and Docker?
Although i dont have direct experience with it because we didnt really have this infrastructure, I understand the functionality and I am eager to learn on the job.
But i understand that SageMaker is a tool for building, training and deploying machine learning models in the cloud
and Docker is used for making sure that machine learning models run the same way across different environments
* AWS SageMaker is a great tool for building, training, and deploying machine learning models in the cloud. It integrates with other AWS services and helps automate things like model tuning, making it easier to manage the entire workflow without worrying about infrastructure. * Docker is really useful for packaging machine learning models and their dependencies into containers. This makes sure they run the same way across different environments, which is key when deploying models in production or moving them between teams.
Can you tell me about a situation where you had to explain a complex concept to someone without a technical background?
I once had to explain NLP model (Roberta Pre trained twitter model) to a non technical stakeholder
I explained that in general NLP is a field in AI that focuses on enabling computers to understand human language.
I then explained that I chose this model because it pays close attention to the relationships between words and their meaning in different context rather than each word separately. For example, it can understand that the word “bank” might mean different things depending on the context, it might mean a financial institution in one context and a river bank in another.
I then explained that I chose this pre-trained twitter submodel in particular because it is trained on massive amount of text from Twitter, which is similar to the forum where I was getting my data from, so it learned patters in how people use language in short, informal posts. It understands abbreviations, hashtags, and slang.
Can you walk me through a recent ML project that you did?
Yes I think a great example is a project that I did at Tesla which was about investigating the impact of conversational AI on the perception of Tesla’s Autopilot and Full_Self Driving products, which are also AI Based
Recently ChatGPT and other conversational AI models suddenly became very accessible to general public and I they became a part of our daily lifes. We as a society never adopted AI to the same extend before, at least not in this form. So it was a significant historical moment and we don’t know how it changes people’s perception of other AI based products and whether we need to adjust our marketing and communication strategy for those products
so my goal was to answer this question
* First of all I needed a lot of data on people’s perceptions so I found a forum of Tesla owners and fans and I the built a custom Python based web scrapper and scrapped the data from this forum * I then cleaned and organized the data, added some additional columns and so on * then I tried a few different NLP models for sentiment analysis and I tested the performance and choose the best one and ran on the whole dataset * then I used a regression analysis to analyze the actual changes in the sentiment before and after popularization of Chat GPT * and I identified the changes in the sentiment and answered the research questions * I then shared it with the product and marketing teams and discussed the implications of my findings
Can you tell me about a project that didnt go as expected?
SkinSafe was supposed to provide data about allergens and population that got tested etc for further use for the optimization of their product scoring system
But in the end they didnt share any data besides the notes from the doctors
Tell me about a time you made a mistake in the analysis and what you did
Example with the dashboard for the competition
Immediately calculated the financial impact and shared it with the leadership and offered solutions
Tell me about a project where you had a change in the requirements and how you approached it
SkinSafe was supposed to provide data about allergens and population that got tested etc for further use for the optimization of their product scoring system
But in the end they didnt share any data besides the notes from the doctors
+ constant changes in the KPIs or shifting priorities
How do you stay up to date?
Im following TLDR Newspaper, they send an update on data scienceand AI topics every day
yesterday for example I learned about the upcoing launch of Apple Intellience (July)
Apple is also launching a conversational AI model, they will integrate it into the existing products
you can now make custom emodji , it will be available in Mail, messages, pages, notifications etc
Siri will be improved
One amusing and relevant AI-related development you could mention is how AI has recently entered the world of professional comedians (sort of). Recently, researchers at Kyoto University developed an AI system that can generate puns by using wordplay algorithms! The AI scours dictionaries and word databases to find words that sound similar but have different meanings, aiming to create funny or amusing sentences. While most of the AI’s jokes are, to put it nicely, pretty bad, it’s an interesting example of how AI is trying to “learn” human creativity.
You could say something like:
“So, the latest thing I learned was about an AI that tries to crack jokes and create puns. It’s still not ready to steal the spotlight at any stand-up comedy clubs, but who knows, maybe one day our virtual assistants will not only remind us of appointments but also make us laugh! For now, it’s mostly dad jokes and bad puns, but it’s a fun step towards human creativity in AI.”
Can you tell me about a project where you had to work with a difficult team member?
Customer Communications team wanted an analysis of their email performance and a life dashboard
but there were a ton of data issues
and it took me a lot of work to investigate where it’s coming from
we figured out that it was a system issue rather than a data replication issues
so we had to involve system project managers
Can you tell me about a time you had to deliver on a tight deadline?
The project for the cost per incremental sale was urgent due to new quarter
it came up and quickly became a priority
So I had to de-prioritize another project that I was working on for the L&D department where we were analyzing the impact of learning on employee performance
