![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230629.395-300x156.png\")
<\/p>\n<\/p>\n
\u2014 Hello everyone and glad to be here and how to design AI for truly intelligent conversations<\/p>\n
\u2014 Goal to move AI from reactive tool to proactive guide, that works with natural and intutitive interaction<\/p>\n
\u2014 Before we deep dive, let\u2019s ask if you have had bad experience with chatbot or virtual assistants<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230631.235-300x151.png\")
<\/p>\n
\u2014 Evolution of conversation design, and see where we came from<\/p>\n
\u2014 Then had intent classifiers, which mapped user references to pre-defined labels and decision trees and recognized variations of requests<\/p>\n
\u2014 Now with LLM have new era of contextual inference of inferring the meaning of the input from the user and grasp intent behind the words<\/p>\n
\u2014 Think of GPS, which used to require the end user to put in exact steps, but it evolved to understand broader instructions and calculate routes dynamically<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230633.341-300x153.png\")
<\/p>\n
\u2014 Building on understanding of conversational design, brought focus to specific challenges team faced to create a truly helpful AI shopping assistant<\/p>\n
\u2014 Assembled team of diverse and dedicated members (conversation designers, data scientists, and researchers)<\/p>\n
\u2014 Important to have design and data working together<\/p>\n
\u2014 Initial investigation of research and analysis showed key problem<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230635.418-300x169.png\")
<\/p>\n
\u2014 There are two layers of user friction<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230637.543-300x169.png\")
<\/p>\n
\u2014 Example of catalog visualization paradox<\/p>\n
\u2014 Efficient browsing and comparison<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230639.485-300x172.png\")
<\/p>\n
\u2014 This approach has drawbacks<\/p>\n
\u2014 How can we maintain benefits of rich bandwidth without overwhelm with personalized guidance and compliments?<\/p>\n
\u2014 Now let\u2019s look at conversational AI interaction<\/p>\n
<\/p>\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230642.030-300x169.png\")
<\/p>\n
\u2014 To solidify this, see the example above<\/p>\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230644.092-300x170.png\")
<\/p>\n
\u2014 Example of expectation gap of how AI should work, with how it actually answers questions<\/p>\n
\u2014 If an AI fails to meet unspoken expectations, a user loses confidence in the system\u2014 its equivalent to someone not answering the questions you ask them<\/em><\/p>\n <\/p>\n \u2014 Our team was faced with mountain of evidence for broken UX, from rage-clicks to a drop-off rate in usage<\/p>\n <\/p>\n \u2014 We analyzed how users interact with the assistant and thousands of chat-logs and phrasing request<\/p>\n \u2014 Three patterns for user queries<\/p>\n \u2014 Traditional bots treated all these queries in similar way and relied on keyword matching, regardless of diversity of input<\/p>\n <\/p>\n \u2014 Concrete example of this can be seen above<\/p>\n \u2014 If asking for gaming laptop it implied they need processor and constraints, like budget available, and preference, a desirable but flexible criteria<\/p>\n <\/p>\n \u2014 Here\u2019s how AI assistant applied framework<\/p>\n \u2014 AI thinks through the conversation and considers what details need to be provided, and crafted prompts with logic to guide assistant to decision making process<\/p>\n \u2014 Architecture is what makes assistant adaptive, and it allows to handle wide range of user inputs, with flexibility and intelligence<\/p>\n <\/p>\n \u2014 Here\u2019s a side-by-side comparison of the current and improved experience<\/p>\n \u2014 In the status quo<\/p>\n \u2014 From the AI assistant<\/p>\n \u2014 This shows how power of structuring AI behavior and guiding intelligence to shape user experience<\/p>\n <\/p>\n \u2014 Favorite part was results we saw with 2.5x increase in monthly active users and using assistant more frequently and more time interacting with it, and higher-level of user satisfaction and engagement and substantial last touch, and taking desired actions with different AI assistants<\/p>\n \u2014 What excited us the most was qualitative feedback from users<\/p>\n \u2014 More natural language, and fewer frustration markers and demonstrate. Power of designing for reasoning and not a response<\/p>\n <\/p>\n \u2014 Arrive at core takeaway of guiding for reasoning as opposed to response, and design for reasoning<\/p>\n <\/p>\n \u2014 Thank you<\/p>\n Q&A<\/span><\/p>\n <\/p>\n","protected":false},"excerpt":{"rendered":" \u2014 Hello everyone and glad to be here and how to design AI for truly intelligent conversations How do we bridge gap between expectations of AI assistant and the actual experience? \u2014 Goal to move AI from reactive tool to proactive guide, that works with natural and intutitive interaction Reasoning is model for structuring information […]<\/p>\n","protected":false},"author":334,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_acf_changed":false,"_seopress_robots_primary_cat":"","_seopress_titles_title":"","_seopress_titles_desc":"","_seopress_robots_index":"","om_disable_all_campaigns":false,"footnotes":""},"class_list":["post-8207","page","type-page","status-publish","hentry"],"acf":[],"_links":{"self":[{"href":"https:\/\/rosenfeldmedia.com\/designing-with-ai\/wp-json\/wp\/v2\/pages\/8207","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/rosenfeldmedia.com\/designing-with-ai\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/rosenfeldmedia.com\/designing-with-ai\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/rosenfeldmedia.com\/designing-with-ai\/wp-json\/wp\/v2\/users\/334"}],"replies":[{"embeddable":true,"href":"https:\/\/rosenfeldmedia.com\/designing-with-ai\/wp-json\/wp\/v2\/comments?post=8207"}],"version-history":[{"count":3,"href":"https:\/\/rosenfeldmedia.com\/designing-with-ai\/wp-json\/wp\/v2\/pages\/8207\/revisions"}],"predecessor-version":[{"id":8319,"href":"https:\/\/rosenfeldmedia.com\/designing-with-ai\/wp-json\/wp\/v2\/pages\/8207\/revisions\/8319"}],"wp:attachment":[{"href":"https:\/\/rosenfeldmedia.com\/designing-with-ai\/wp-json\/wp\/v2\/media?parent=8207"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230646.057-300x169.png\")
<\/p>\n\n
\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230648.343-300x169.png\")
<\/p>\n\n
\n
\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230650.851-300x168.png\")
<\/p>\n\n
\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230652.837-300x169.png\")
<\/p>\n\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230654.973-300x160.png\")
<\/p>\n\n
\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230657.097-300x172.png\")
<\/p>\n\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230659.146-300x173.png\")
<\/p>\n\n
\n
![\"\"](\"https:\/\/rm-media.nyc3.digitaloceanspaces.com\/wp-content\/uploads\/sites\/36\/2025\/06\/image-2025-06-10T230701.508-300x169.png\")
<\/p>\n\n
\n
\n
\n
\n