United Nations Chatbot: Native Mobile Application Prototype

Lauren Beronja

Timeline — About two weeks

Workforce — Lauren Beronja, Kacy Yang

Function — UX / UI Designer, Researcher, Data Architect


The UN Is on the lookout for a cell chatbot and useful resource software. The applying could be downloaded by group members who search help across the workplace’s 9 classes of points. The consumer might want to have entry to the appliance on iOS and Android, anyplace on this planet, with an choice to obtain or entry the knowledge offline.

We kicked off our challenge with a analysis section, which primarily consists of two parts- mapping points to useful resource supplies, offered by shoppers and on-line, and have analysis by way of comparative evaluation. Via useful resource materials analysis, we found a listing of ache factors lists as follows:

Mapping out points to locations

We discovered that there’s an informational hole between two key sources: The primary points and the respective places of work. The sources are situated in several recordsdata which are full of content material which makes it onerous to navigate. The primary ache level that we discovered is that it could be troublesome for workers which have questions or points discover the respective workplace in an environment friendly means.

Workplaces info must be organized higher

We additionally found that the knowledge within the workplace listing file “The place to go when”, although informational, shouldn’t be very clearly structured particularly for individuals who have points and are in search of assist. The file requires the customers to have a considerably good understanding of every workplace’s position and/or what they do. The construction of data is counter intuitive when a consumer has points and is trying to establish which workplace that they need to contact.

Comparative Evaluation

Via comparative evaluation of quite a few Chatbot Mobile Purposes we got here up with a options that we believed could be needed in our Chatbot prototype.

  • A majority of the cell functions available on the market supply their service by way of each iOS and android platforms.
  • About 60% of the cell apps which are in the same style have studying associated modules. We discovered that studying modules can improve info digestion which ends up in a extra environment friendly downside fixing course of.
  • We found that the typical age of workers within the group ranges from 42–49. Via unofficial consumer interviews, we discovered {that a} vital quantity of customers in that demographic would like a standard catalog or listing fashion search system, versus a chatbot.
  • Chatbots are useful however not excellent. Providing a stay chat characteristic is critical particularly for pressing conditions.

Localization and accessibility

Having workers members working in lots of international locations around the globe implies that the Application should embody choices to allow the consumer to alter the language.

Moreover, we perceive that the corporate hires many workers members which will have a necessity for accessibility options, notably round imaginative and prescient and listening to impairments.

Chatbot and Listing — the 2 means method

With the ability to search a listing is a vital characteristic that gives clear and structured info for customers. After analysis, we determined to design our software with a two means approach- an app that has chatbot as a major listing and information, and a standard listing as a secondary assist characteristic. Our design aim is to get customers to search out info effectively by way of a number of means.

Data structure

We organized given info into three major classes: Points, Organizations, and Processes:

Points will lead the movement with subject identification, adopted by detailed info and sources with respective places of work’ contact info.

The organizational listing features in a standard way- it begins with workplace description which leads customers to further sources within the respective discipline, points, and speak to info.

Within the course of web page, info of the casual course of and the formal course of are provided. We additionally determined to incorporate a key timeline portal for fast and direct entry.

Usability testing was essential to us all through our design course of. We have been in a position to obtain suggestions throughout the usability checks for each our Mobile Application Prototype in addition to the Chatbot which aided us in making the wanted iterations of each.

For the Mobile Application we carried out six usability checks.

Suggestions included:

  • Fixing interplay bugs (e.g damaged scrolling perform)
  • Altering our menu classes from “Assets” to “Concern Identification” to keep away from confusion
  • Embody a “Visitor log-in” characteristic to offer ease of mine for individuals who desire to really feel utterly nameless.

For the Chatbot we carried out 4 usability checks.

Suggestions included:

  • AI chat perform enchancment with pure speech processing
  • Extra key phrases to assist additional establish consumer’s points
  • Easy out the dialog logic and loops to make it mimic pure dialog.

Chatfuel is a chatbot constructing platform designed for messenger. In our case, we used Chatfuel to construct our AI enabled chatbot “Toby” for demonstration functions. We began with mapping out the chat movement logic with a chatbot consumer movement diagram, then moved on to constructing out key chat scripts earlier than we began constructing in Chatfuel.

Our chatbot’s characteristic consists of:

  • AI pure dialog and fundamental speech/key phrase identification.
  • Points listing
  • Workplace listing
  • Course of listing
  • Dwell chat assist for pressing circumstances
  • Expandable studying modules
  • Different chatbot fundamental features

Chatbot Prototype Demo hyperlink: https://www.messenger.com/t/102807467817597

We began implementing our findings by determining what the consumer movement or website map of the cell software would possibly embody.

Chatbot App constructed with Axure and Figma

We used each Figma and Axure to construct out our chatbot app prototype. Figma was used as a way to shortly create low-fidelity wireframes collaboratively. We then converted to Axure to construct the high-fidelity prototype and embody interactive options in each iOS and Android. Another design instruments are used alongside the method, resembling Adobe Photoshop and Illustrator for graphics and icons.

After analysis and have prioritization, our chatbot cell software has some key options as follows:

Streamlined Onboarding

The Onboarding is nameless and customizable in order that the consumer is aware of their info is not going to be shared and their points will likely be solved in a faster method. We additionally included the stay assist directions in order that the consumer will know they’ll chat in actual time in case of an emergency.

Dwell Help Choice for Emergencies

Dwell Help possibility for rapid service and a Chatbot — because the highlight characteristic — to offer custom-made steerage for individuals who are in search of assist

Twin Listing System

The consumer is ready to entry info by way of two directories. The primary listing is organized by sure points the consumer could also be having whereas the second listing is organized by workplace.

Localization and Accessibility choices for inclusive design

As a way to assist the Shopper with future design and construct out of the Mobile Application Prototype we created a roadmap and categorized it in 4 phases. Though we didn’t get a lot details about the group of builders and designers that will be engaged on the Application, have been have been in a position to give you a basic thought of what every section would appear to be — not together with a timeline.

We categorized our high-level roadmap in 4 phases:

The prototype section, the MVP growth section, the Alpha section and the Beta section.


Please enter your comment!
Please enter your name here