Approach:

1. Understand the prompt better and establish the bussiness goal

2. Design a new app

3. Define user segment and pain points and solutions

Understand the prompt better and establish the bussiness goal

Candidate: Do we need build new app to make use of this api or can we use the existing app for this api

Interviewer: New app

Candidate: Is it a paid app or free app?

Interviewer: You decide

Candidate: Are we trying to leverage this new technology to serve the google users better and boost the company’s topline

Interviewer: Yes

Potential app:

Mind reader: Users can use the app to others currrent state of mind

User segments:

1) Organization that recruit people

2) People that want to choose the life partner

3) Sales people that want to tailor their sales pitch according to the audience mood

I would prioritize the user group that has a biggest customer base. Most of the people want to choose a life partner, therefore I would select this group. Also by facilitating long-term marriage, we help to create stable relationship, healthy families and properous communities.

Pain points:

A) Difficult to identify potential romantic partner

B) It takes long time and lot of energy to understand the compatibility with the potential life partner

C) Divorces/break-up happens because people misunerstand the other person

Solution:

A) Difficult to identify potential romantic partner:

With the new app, users can learn about the other person from the psyiological response of the other person. For example, if the other get excited and become happy, then that could be signal that the other person is interested in a romantic relationship.

Effort: high, need to develop a new app from ground up to read the information from the api

Impact: high, no other app in this work serves this use case today, so we can create a unique value proposition

B) Takes time to test the compatibility:

App would develop a perfect match model based on the historical data of the past successful marriage. App would suggest a set of questions to ask the potential partner and set of activities to do with the potential partner and the app would track portion of time they both were happy, content, comfortable from their mood signals and provide a compability score.

Higher a score, higer the compability.

People can know about their compability with other within a day rather than years

Effort: Moderate, need to build a personalization engine to come up with a set of questions and activities and ML model to develop a matching algorithm

Impact: High, will build a stable marriages

C) Divorces happen becuase people misunderstand the other person:

Some people are not great at articulating their thoughts or not very diplomatic but they might be well meaning.

For example, people might say something in jest or is being sarcastic, otherwise person might mistake, leading to strain in the marriage

In such instances, when there is a conflict between the words and underlying intentions/moods of the partner, this app will send the pust notification to the user, leading to better understanding of the other person and healthy relationships

Effort: low, need to build a rule-based system to identify the cases where words do not match the emotions, and notification system to alert the user about such instances

Prioritization 

If I were to prioritize on one solution, I would choose initiative to test the compatible partners. Becuase more compatible the partners are, more happy their marriages be, more happy the entire society become.

To begin with I would like to list down.
1) What sort of apps are possible?
2) I will select some which are relevant for Google
3) Recommend one App that we should build and why?
4) Talk about the success metrics for that app.
5) Summerize what we did so far.
—————-
1) What sort of apps are possible.
A) Social media apps can use this but who want the world to know if I am sad.

But you could have permissions like your close friends and family can know your emotional
state.

B) Counseling app
C)  Alexa, GoogleHome kind of app – I will port the API to them
D) AI-enabled Companion pet/friend
E) I will integrate into car driving systems
F) Online Meeting app can use this to let other participants know hat someone is stressed out,
feeling tired. of seemingly looking confused.
G) Live online learning platforms can use it to know when students are confused.
H) recorded a video learning platform like Coursera/Uedmy use it to know which portion of their video people were most confused.
——————

2) Among all of these if I select which of these are relevant for Google.

A) Social media use case is relevant as Google has Google Plus but more than often people don’t want to let the worlds know when they are sad, stressed. In fact, people often fake on social media that they are the happiest person on the planet.

Yes, people can select to show emotions only to very close friends and family. But again in order for this feature to be successful people need to give consent, and in today’s world kids are hesitant to accept parents and grandparents as friends on social media.  So I suspect they will easily give consent to parents. Yes some of them may happily give consent to share their emotion with their close friends, classmates, etc.

Overall this can definitely have some positive impact on the mental well-being of people but for now, I am moving on to the next solution.

C & D ) is relevant for GoogleHome, but given that API is currently available only for I will keep it for google’s long term research goals. The moon shots.

E) is very relevant to Google Maps:  If I am driving google maps can detect that I am not talking to a co-passenger and just listening to some music. I may get tired my lateness levels are gone down,

it could alert me or why just alert with me rather converse with me, talk to me on topics like good things on last vacation, funny things that happened during the last long drive. keep me engaged so that that I don’t fall asleep and I always have enough alertness.

Let’s call just alerting as v1 and conversation as v2.

F) Google has online meetings, Google Meets, it’s very much applicable there for the effectiveness of the meeting

Idea	User Impact	Ease of implementation
Social media	Medium (Mental well being.)	easy
C & D	High (mental well being + a companion)	very high
E) (Google Maps Navigation)	Very High ( Can save lives)	v1 – easy v2- medium
F) Online Meetings	Medium (Can increase the effectiveness of meetings) I could defend why this is not high and why it’s medium	easy

—-
Recommendation:  I would recommend the Google Maps Navigation idea because the impact is very high.  I will also put the online meeting idea at the very top of my backlog.

OR maybe since given that Google Maps and Google Meet are two different teams. I can ask both ideas to be implemented.

—
Success metrics:
Google Map Navigations: 
North star metrics:  Numer of times emotion detector kicked in and alerted/conversed.

There can be many secondary metrics like every time the emotion detector kicked in ask the user for feedback if that helped, whats their feedback.

GoogleMeet: 
How many times emotion detection kicked in and was correct, ask the presenter for the feedback did emotion detector helped you?
—–
Summary: So overall what we did was
1) We saw, What sort of apps are possible with such API?
2) We shortlisted that make sense for Google
3) We prioritize that and came up with the recommendation that Google Maps and GoogleMeet both should use this.
4) We talked about the success metrics for that app.