Searching for the perfect item while shopping online can be complicated and finding style inspiration can prove to be a time consuming task.
According to our user study, in addition to slow delivery times, 61% of online apparel shoppers reported counterintuitive navigation and filtering as one of the main pitfalls on e-commerce sites.
This is where we stepped in and transformed conventional online shopping into a personalized, interactive experience which reduces time and effort spent on the search process.
What it does
Zally navigates in Zalando and filters items quickly and easily using voice commands.
To do this, it listens to commands and recognizes item categories, colors, sizes, and descriptions of items and browses the Zalando store in efforts of finding the perfect goods, whether it's a new pair of jeans in just the right size or a top to match with items from previous orders. Zally will also guide to the right direction in case of customer service issues or other related queries.
How we built it
SpeechRecognition API is used to record and understand the user's verbal input. With pattern matching through regular expressions, the phrases are parsed and relevant information is extracted and used as basis for filtering.
Zalando article API is then used to get an up-to-date list of clothing items and accessories that match the filters.
Challenges we ran into
The speech recognition built into the browser can occasionally return results which might sound similar to user's request but don't match any actual patterns. As indication of rerunning the query, an error handler is added.