Search and Personalization team's mission in Tiki is to help customers discover products as easy as possible through providing relevant search results and useful recommendation within customer’s browsing context.
To serve this important mission, our team is constantly iterating and standing together to solve problems. We build and maintain features that have extremely high throughput with millions of requests per day and high contribution to company's revenue. We play with Big Data, Machine Learning, and even Deep Learning.
As a Back-end developer working within Search and Personalization team at Tiki, you'll be the key driver for:
- Drive and implement technical solution to a variety of features that serve high throughput and also high availability.
- Build, test and ship back-end APIs support search and recommendation features.
- Design, implement and maintain streaming flow of products, categories, brands, even customer interactions from multiple data sources with high accuracy.
- Understanding business objectives, working closely with Data scientist to analyze, develop, ship and optimize machine learning models used in search ranking and recommendation.
5 reasons why you should join us:
- We are constantly iterating! There is no such best version for anything, no fastest API, no best machine learning models. We build, test, ship, and optimize, and test. Just a stream of improvements and tests.
- We have data-driven mindset, every point of changes must be tested to gain insights into its impacts on key metrics. It's a long process, but over time, we gradually learn and become confident in our approach.
- We have "enough" valuable data to analyze and gain insights into customers, which helps us adjust our features to bring the best experience to customers.
- We love "best practices". Serving important features with high throughput always give us a hitch to research and apply best practices. Any experiment or optimization is always welcomed.
- We are both independent and open. We own our products. Technical problems would be discussed internal, but for difficult one, we could request other's help.