Yecid Sanmartin - Senior Product Designer





Alexa Enabled Search


UX/Interaction Design


Jan 2021 – Oct 2021

Customer Problem: Given the modality of typing, customers were sending 1-2 word requests to Alexa (eg. “settings”, “bluetooth”, “music settings”). Customers expected Alexa to help them to complete their given task, and instead received frictive/dead end responses back (eg. “Hmm, I don’t know that one”, “Sorry, I’m not sure”) resulting in a poor customer experience.

Design Solution: Expand Alexa’s ability to respond to typed requests within a conversation by placing suggested actions/utterances/links in the chat, increasing the likelihood that the customer is able to select a suggested result to complete their task.

– Description

Picked up search project on quick turn from designer who was leaving and quickly delivered conversational and conventional search cx concepts. Requirements unclear from the start and misalignment between product and leadership on direction and problem to solve.

Worked backwards from customers to identify main points of friction within the type with Alexa experience we needed to solve. This helped reduce the complexity of many earlier explorations that included features such as search result categories, domain grouping, tags, and rich media results. Customers simply needed to find what they were looking for in the app to complete a task and expected Alexa to be able to help. Highlights of explorations included 1) app wide ingress 2) components for search suggestions and filtering 3) visualization of results (cards, linking, grouping, media) 4) in-chat controls for devices and controlling media.

Worked with research to run user studies and apply learnings to design iterations and prototypes on tight schedule. Informed product requirements per outcome of user studies and designs. Collaborated with partner teams to explore a conversational search cx for their domain (skills, devices). Delivered Interaction Design Specs for conversational search CX launched Oct 2021.