Design a conversational interface

The problem

In the manufacturing and service sectors, access and optimal data sharing between internal resources requires a considerable expenditure of time and has a significant impact on production costs: the artificial intelligence with tools such as chatbots, rich media cards and semantic databases can represent a valid resource to optimize working processes.

For a pharmaceutical recruiting company, the project goal was to find a solution shared by all parties involved – project managers, human resources and sales – to facilitate the retrieval and matching of consultants curricula and their correct allocation at the client’s facility, taking advantage of the AI.


Keethings (Italy)


  • Lean UX workshop
  • UI Design
  • Prototype


Sketch  Invision

Approach & solution

To accurately identify the customer’s requirements, a single day Lean UX workshop was organized involving stakeholders and product managers. The goal was to create a functional prototype so to satisfy the needs of the primary user.

The workshop was divided into six phases:

  1. Define business objectives
  2. Specify the context of use
  3. Identify the user profile
  4. Describe user scenarios
  5. Provide ideas to the problem
  6. Find a shared solution

Business goals

Each participant highlighted the main objectives to be achieved so to use them as performance indicators (kpi). Eventually, the three major were selected in a priority scale.

Context of use and user profile

The team was asked to analyze business goals from a different perspective, answering a series of questions such as:

  • Who is the user?
  • What problems must the product / service solve?
  • How, where, when is it used?
  • What features are important?
  • What needs must meet?
  • What solutions can we give to these needs?

The information thus gathered was grouped and used to identify a user profile (primary persona) based on demographic characteristics, needs, problems and goals.


Analyzing the user needs and problems, a series of questions were formulated to be translated into scenarios that the team then selected for the next co-design phase.


Participants were asked to provide as many ideas / solutions to user stories in a divergent approach. Each idea was viewed and evaluated by the team to produce a shared solution.


The solution identified was illustrated in a user flow so as to provide a blueprint for a high definition prototype to be validated inside the team.

Key points

  • A welcome rich media card allows to quickly access the main features customized on the user profile: responding to pending requests, viewing the resource allocation status, inserting a new request or consulting the curricula.
  • The chatbot acts as a virtual assistant so to allow the user to access a more detailed and personalized information.
  • Rich media cards and the semantic database allow to generate the most suitable profiles and to filter them according to client’s requests.