Data Science is often labeled as one of the sexiest jobs of the 21st century. But it is really hard to find the right sexy data science job.
More and more companies are trying to collect tons of data on pretty much anything they do and try to hire data scientists in hopes of creating value from this data. But companies are all at different stages of data science capabilities and have different expectations of what data scientists should do. So it is very hard for data scientists to choose the right company for their ideal job.
The data scientist role is used so broadly nowadays that it is very hard to know exactly what the job functions are without detailed descriptions. The data science spectrum can range from one end of performing analytics for insights and defining/dashboarding metrics to the other end of productionizing predictive models to be used at real time. Any person who works on a portion or entirety of these are being widely called data scientists.
I think the most exciting part of being a data scientist is being able to do the entire spectrum of full stack data science. I want to be able to: 1) dig into data to identify areas of opportunity, 2) design and prototype some sort of solution such as predictive, optimization, or even heuristic models, 3) put my model into production and A/B test it to measure the impact, and 4) iterate through the model to improve it. But it is extremely difficult to find such an opportunity and get enough support to be able to do all these.
On top of all these already tough requirements, I wanted to work for a company that is building a product that I truly believe in as a customer. After some searching for my ideal job, I ultimately decided to continue my journey as a data scientist at Thumbtack and here is why:
1. Company Mission
Thumbtack is a two sided marketplace where customers can find professionals for pretty much any job that they can imagine. For example, customers can find house cleaners, DJs, piano tutors, or even traffic law attorneys. It not only makes customers’ lives easier and more exciting, but also provides a living for the professionals; it’s more than just making life more convenient. Many people worry about the AI revolution that would potentially result in loss of many jobs. But Thumbtack, on the other hand, is creating a platform that can help people find jobs beyond driving or delivering food. It is enabling people to do what they are good at and actually love doing. Not only is this meaningful but this is a $700B market that has not been cracked yet. There is a tremendous opportunity for growth with Thumbtack leading the charge.
2. Data & Projects
As a data scientist, you want to be able to work with all sorts of data. Thumbtack provides a platform where people browse and choose the professionals they want for their jobs, write messages and have conversations about the project details, and even post pictures or videos related to the job. This results in all sorts of fun behavior data, text data for NLP, and pictures / videos for computer vision. These are not just possible data Thumbtack can have, but are data that already accessible with the state-of-art data warehouses.
With this data, Thumbtack is in the process of tackling many exciting projects in many different parts of the company. Thumbtack is trying to minimize risk for both customers and pros through projects such as spam detection, improve monetization through user intent models, optimize customer experience by developing ranking models, and even manage user generated images through object detection. These are just the tip of the iceberg of problems Thumbtack wants to take on. The Data Science team aims to be pretty much everywhere in the company to build technologies that serve as strategic differentiators.
3. Team Structure
Two of the most important things I was looking at was: 1) does this company actually value data science?, and 2) how is their data infrastructure? One way to check these are to look at how they’ve structured their teams and what they have set up.
In many companies, data scientists are expected to be jack of all trades and they end up working on data collection, modeling, dashboarding, and ad hoc analytics reports. All of these are very important and things data scientists should consider, but the time load is usually very unbalanced such that most of the time is spent on data pipelining, dashboarding, and ad hoc reporting. This often limits data scientists from building impactful products that they are truly excited about. I’ve seen companies promise that data scientists can spend 10~20% of the time solving problems or developing models, which is too low from my expectation.
But Thumbtack was very impressive on this end. Thumbtack has a Technical Infrastructure team that manages infrastructure that can collect the right data with the optimal method. Thumbtack has Modeling Platform team that sets up tech like TensorFlow to ensures data scientists can deploy their models to production with ease. There is a separate Data Analytics team that helps other stakeholders understand data and manage dashboards to condense data into insights. This all showed that Thumbtack really makes a joined effort to extract value from data and that the Data Science team can really focus on the “data science” part of the work to design/build/optimize problems.
These were all things I considered as an outsider of the company. But having joined the company, I love my decision even more. All the expectations are held true, the culture is great, the leadership team is super transparent, the company direction is focused, and more. I get to work on cool projects such as professionals risk profiling to make Thumbtack a safe platform for the customers. I have opportunities to get involved in projects such as search ranking to improve the quality of list of professionals we show to the customers. Also,I am part of text modeling project for understanding conversations between customers and professionals. If all of these sound exciting to you as well, come join Thumbtack Data Science!