Meet Luna

- Designing for Conversations

 

THE OBJECTIVE 

To improve the experience of scheduling meetings in the enterprise through a 6-week exploration. 

THE OUTCOME

The interface for a cognitive conversational assistant within the enterprise which cleared funding for 2 RSMs (Research Staff Members) in a joint development model to start implementation.

 

 

 

 


Meet Luna

- Designing for Conversations

 

THE OBJECTIVE 

To improve the experience of scheduling meetings in the enterprise through a 6-week exploration. 

THE OUTCOME

The interface for a cognitive conversational assistant within the enterprise which cleared funding for 2 RSMs (Research Staff Members) in a joint development model to start implementation. 

 

 

download (21)

 

THE STORY.


As part of on-boarding to IBM Design, the studio runs an incubator program, which takes some of the hardest problems within various units of IBM and puts design as a focus to create a solution. Together, all disciplines of designers and product managers create solutions which culminate in a deliverable to business unit executives.

We were given 6 weeks to explore the possibility of a cognitive conversational assistant which could improve scheduling within the enterprise. Our process began with examining the overall space of scheduling and issues faced by employees - both internal and external to IBM - and then diving into extensive testing of our concepts and ideas.

 

 

THE STORY.


As part of on-boarding to IBM Design, the studio runs an incubator program, which takes some of the hardest problems within various units of IBM and puts design as a focus to create a solution. Together, all disciplines of designers and product managers create solutions which culminate in a deliverable to business unit executives.

We were given 6 weeks to explore the possibility of a cognitive conversational assistant which could improve scheduling within the enterprise. Our process began with examining the overall space of scheduling and issues faced by employees - both internal and external to IBM - and then diving into extensive testing of our concepts and ideas.

DISCOVERY AND RESEARCH.

The first phase of our process focused on discovering the problem at hand and identifying the key users. We dove right into generative research to understand the real problems behind scheduling.

Rapid + In-Depth Interviews: We conducted a sum total of 46 interviews; both in-depth and rapid with a broad set of users. Once synthesizing all the data and creating artifacts such as empathy maps, we were able to construct two personas.

DISCOVERY AND RESEARCH.

The first phase of our process focused on discovering the problem at hand and identifying the key users. We dove right into generative research to understand the real problems behind scheduling.

Rapid + In-Depth Interviews: We conducted a sum total of 46 interviews; both in-depth and rapid with a broad set of users. Once synthesizing all the data and creating artifacts such as empathy maps, we were able to construct two personas.

 

 

MAKING SENSE OF IT ALL.

Once we had narrowed down who our primary users were, we could begin to identify and synthesize their main pain points that we wanted to design for. Through this, we were able to segway right into ideation and begin thinking of the possibilities that we could start to design.


We quickly decided that we wanted to first not limit ourselves, and think of as many 'big ideas' and 'analogies' as we could. We unleashed our ultimate creativity and worked on generating and mashing up various ideas. Our principle was to make as much as we could and subsequently test as fast as we could.
 

 

 

MAKING SENSE OF IT ALL.

Once we had narrowed down who our primary users were, we could begin to identify and synthesize their main pain points that we wanted to design for. Through this, we were able to segway right into ideation and begin thinking of the possibilities that we could start to design.


We quickly decided that we wanted to first not limit ourselves, and think of as many 'big ideas' and 'analogies' as we could. We unleashed our ultimate creativity and worked on generating and mashing up various ideas. Our principle was to make as much as we could and subsequently test as fast as we could.
 

 

 

MAKING SENSE OF IT ALL.

Once we had narrowed down who our primary users were, we could begin to identify and synthesize their main pain points that we wanted to design for. Through this, we were able to segway right into ideation and begin thinking of the possibilities that we could start to design.


We quickly decided that we wanted to first not limit ourselves, and think of as many 'big ideas' and 'analogies' as we could. We unleashed our ultimate creativity and worked on generating and mashing up various ideas. Our principle was to make as much as we could and subsequently test as fast as we could.
 

 

Luna Images.001
Luna Images.002

 

 

All members of our team were roaming the hallways, finding as many people that we could run our ideas by as we could. 

We made quick and dirty prototypes and worked on first testing the overall reception of our ideas and concepts, versus focusing on the general usability. By thinking of as many ideas as we could, and testing them immediately, we were able to keep our creative juices flowing and also move towards creating a single, unified solution. By enabling all members of the team to participate, we were able to accomplish a great deal of ideation and testing in a very short amount of time. We truly divided and conquered.

 

 


All members of our team were roaming the hallways, finding as many people that we could run our ideas by as we could. 

We made quick and dirty prototypes and worked on first testing the overall reception of our ideas and concepts, versus focusing on the general usability. By thinking of as many ideas as we could, and testing them immediately, we were able to keep our creative juices flowing and also move towards creating a single, unified solution. By enabling all members of the team to participate, we were able to accomplish a great deal of ideation and testing in a very short amount of time. We truly divided and conquered.

 

 

 

MEET LUNA.

Luna is a cognitive conversational assistant to help users re-define the terrible experience of scheduling meetings. Luna can be integrated with a variety of portals, including enterprise e-mail, Slack, and even your phone - Luna meets users where they already are.  

Luna Images 2.001
xx.001
Luna Images 2.003

 

 

TEACHING LUNA.

While designing Luna's interface and core functions was important, we also realized that we must think through Luna as a learning model. What would be the various steps Luna would take in order to learn and adapt to users? Moreover, how would we measure success? 

First, we must teach Luna the basics. Luna would be given access to the user's calendar and scheduling history through Verse (e-mail) and Slack integration. Luna must be able to speak and understand basic natural language.

Second, Luna's learning algorithms will get better from more data (e.g. behavior, patterns). Luna will also make minor suggestions, which feeds into the data collection based on user acceptance or rejection. Luna will also have more permissions, such as contacting participants in users' meetings, creating user groups etc.

Finally, once the learning algorithms have been made stronger, Luna can become more proactive, recognizing past meetings and making predictions without user initiation. Further, Luna can integrate with the communication platforms users want to use. 


 

 

 

TEACHING LUNA.

While designing Luna's interface and core functions was important, we also realized that we must think through Luna as a learning model. What would be the various steps Luna would take in order to learn and adapt to users? Moreover, how would we measure success? 

First, we must teach Luna the basics. Luna would be given access to the user's calendar and scheduling history through Verse (e-mail) and Slack integration. Luna must be able to speak and understand basic natural language.

Second, Luna's learning algorithms will get better from more data (e.g. behavior, patterns). Luna will also make minor suggestions, which feeds into the data collection based on user acceptance or rejection. Luna will also have more permissions, such as contacting participants in users' meetings, creating user groups etc.

Finally, once the learning algorithms have been made stronger, Luna can become more proactive, recognizing past meetings and making predictions without user initiation. Further, Luna can integrate with the communication platforms users want to use. 


 

 

 

TEACHING LUNA.

While designing Luna's interface and core functions was important, we also realized that we must think through Luna as a learning model. What would be the various steps Luna would take in order to learn and adapt to users? Moreover, how would we measure success? 

First, we must teach Luna the basics. Luna would be given access to the user's calendar and scheduling history through Verse (e-mail) and Slack integration. Luna must be able to speak and understand basic natural language.

Second, Luna's learning algorithms will get better from more data (e.g. behavior, patterns). Luna will also make minor suggestions, which feeds into the data collection based on user acceptance or rejection. Luna will also have more permissions, such as contacting participants in users' meetings, creating user groups etc.

Finally, once the learning algorithms have been made stronger, Luna can become more proactive, recognizing past meetings and making predictions without user initiation. Further, Luna can integrate with the communication platforms users want to use. 


 

 

 

TEACHING LUNA.

While designing Luna's interface and core functions was important, we also realized that we must think through Luna as a learning model. What would be the various steps Luna would take in order to learn and adapt to users? Moreover, how would we measure success? 

First, we must teach Luna the basics. Luna would be given access to the user's calendar and scheduling history through Verse (e-mail) and Slack integration. Luna must be able to speak and understand basic natural language.

Second, Luna's learning algorithms will get better from more data (e.g. behavior, patterns). Luna will also make minor suggestions, which feeds into the data collection based on user acceptance or rejection. Luna will also have more permissions, such as contacting participants in users' meetings, creating user groups etc.

Finally, once the learning algorithms have been made stronger, Luna can become more proactive, recognizing past meetings and making predictions without user initiation. Further, Luna can integrate with the communication platforms users want to use. 


 

 

TEACHING LUNA.

While designing Luna's interface and core functions was important, we also realized that we must think through Luna as a learning model. What would be the various steps Luna would take in order to learn and adapt to users? Moreover, how would we measure success? 

First, we must teach Luna the basics. Luna would be given access to the user's calendar and scheduling history through Verse (e-mail) and Slack integration. Luna must be able to speak and understand basic natural language.

Second, Luna's learning algorithms will get better from more data (e.g. behavior, patterns). Luna will also make minor suggestions, which feeds into the data collection based on user acceptance or rejection. Luna will also have more permissions, such as contacting participants in users' meetings, creating user groups etc.

Finally, once the learning algorithms have been made stronger, Luna can become more proactive, recognizing past meetings and making predictions without user initiation. Further, Luna can integrate with the communication platforms users want to use. 


 

 

...

Project Luna was a wonderful experience for me to kick-off my time at IBM Design. It was the first enterprise project I was able to lead in terms of research, and I was fortunate to have the stellar team that I did with me. I quickly learned the importance of knowing how much research is enough, and how to use various methods to keep moving forward on a project. I was also able to learn the intricacies of conversational design by collaborating with my wonderful UX designers. I can't wait till Luna is available for us mortals - we all know we need a Luna in our lives! 

 

 

 

...

Project Luna was a wonderful experience for me to kick-off my time at IBM Design. It was the first enterprise project I was able to lead in terms of research, and I was fortunate to have the stellar team that I did with me. I quickly learned the importance of knowing how much research is enough, and how to use various methods to keep moving forward on a project. I was also able to learn the intricacies of conversational design by collaborating with my wonderful UX designers. I can't wait till Luna is available for us mortals - we all know we need a Luna in our lives! 

 

 

 ...

Project Luna was a wonderful experience for me to kick-off my time at IBM Design. It was the first enterprise project I was able to lead in terms of research, and I was fortunate to have the stellar team that I did with me. I quickly learned the importance of knowing how much research is enough, and how to use various methods to keep moving forward on a project. I was also able to learn the intricacies of conversational design by collaborating with my wonderful UX designers. I can't wait till Luna is available for us mortals - we all know we need a Luna in our lives! 

 

 

 

TEAM

KATE BLAIR - LEAD

RUTVI GUPTA - RESEARCH

HANNAH MOYERS - UX

REBECCA LOVE - UX

KYLEE BARNARD - VISUAL

ROBERT THELEN - OFFERING

LANA LI - OFFERING


 

TEAM

KATE BLAIR - LEAD

RUTVI GUPTA - RESEARCH

HANNAH MOYERS - UX

REBECCA LOVE - UX

KYLEE BARNARD - VISUAL

ROBERT THELEN - OFFERING

LANA LI - OFFERING


 

TEAM

KATE BLAIR - LEAD

RUTVI GUPTA - RESEARCH

HANNAH MOYERS - UX

REBECCA LOVE - UX

KYLEE BARNARD - VISUAL

ROBERT THELEN - OFFERING

LANA LI - OFFERING

 

 

SKILLS

DESIGN RESEARCH

STRATEGY

UX DESIGN

DESIGN THINKING

 

 

 

ROLE

RESEARCH

IDEATION

STRATEGY

PRESENTATION