Carpe DM.

A mental health application for young people which uses machine learning and conversational UI to help them keep on top of their daily struggles. Carpe DM is unique in that it's personal assitant CaseyBot adapts to the user, resulting in a companion that is more enaging, helpful and personal.

  • Role
  • Lead Designer
  • Year
  • 2017
  • Company
  • Personal
  • Tools
  • Sketch / Invison
  • Platform
  • Mobile App

The Problem.

In a 2013 report, The Royal College of Surgeons (RCSI) found that by the age of 13, 1 in 3 young people in Ireland are likely to have experienced some type of mental disorder. By the age of 24, that rate will have increased to over 1 in 2. Based on international evidence, RCSI believes that this means that up to one third of young Irish adolescents and over one half of young Irish adults are at increased risk of mental ill-health. These results clearly show that mental health is an epidemic that needs to be addressed both in Ireland and around the world.

An examination of the population of Ireland’s mental health, on a county by county basis, was conducted using the five item Mental Health Index (MHI), a subscale of the widely used SF-36 (Short Form Health Survey) (Houghton, Keane, Murphy, & Dunne, 2011). The MHI works on a scoring system out of 100 (100 being the most mentally fulfilled), and has independently been found to be a quick, accessible, valid and reliable marker of mental health (McCabe et al., 1996). The study found that the mean score of the population of Dublin, the most urban county in Ireland, was 79.8, whilst the mean score in Limerick, one of the most rural counties in Ireland, was 85.2. These findings show that on average Dublin has more mental health sufferers than Limerick. Similar results are visible in studies carried out across the world, leading to the belief that people are more likely to suffer from mental health issues in urban communities than they are in rural communities.

That being said however, the National Suicide Research Foundation (2014) reported that there are around 11,000 episodes of deliberate self-harm presenting at A&E departures each year and up to 500 suicide deaths reported in Ireland. Out of those 500 cases of suicide, the majority take place in rural areas. The HSE (2015) released the 3-year moving average rates of suicide in Ireland by geographical area. This report highlighted that the three areas most bereaved with suicide are Limerick (22.4 / per 100k), Carlow (17.5 / per 100k), and Roscommon (16.2 / per 100k). Urban areas such as Dublin City came in at 7.9 per 100k and Cork city came in at 12.1 per 100k. Fontanella (2014) found startlingly similar results in the United States, commenting that across her study rural suicide rates were nearly double those of urban areas for both males (19.3 and 10.31 per 100k, respectively) and females (4.40 and 2.39 per 100k, respectively).

Of the 500 total deaths by suicide per year, nearly 400 of these deaths are male (19.1 per 100k compared to 4.5 per 100k female). The age group with the most suicides recorded is the 18 – 35 bracket. Based on this it is clear that the highest risk population for suicide are young males living in rural communities. Dr. Sandra Micken (2015) has summarised why this particular group is so susceptible to suicide as being; lack of access to mental health services, social isolation, and a “macho” mindset where men feel they cannot openly speak about their problems.


Create an aesthetically pleasing product which applies psychology principles in order to improve the day to day life of people in rural communities. Obviously this is a huge challenge which could not possibly be solved in six weeks, whilst trying to cover the requirements of a college brief, however it was a particularly exciting and engaging proposition.

The Solution.

The competitor analysis conducted at the beginning of this project identified a number of applications which were suitable for redesign and redevelopment. However it was my belief that these applications were too light on features and did meet the needs or requirements of those identified during the initial research phase. Because of this Carpe DM will be created from the ground up as an application which takes on ideologies of some of the more popular apps as well as new features that are linked to the key findings on the topic of mental health.

Carpe DM, at its core, is a to-do like application which will allow users to keep track of their daily habits and plan in advance, helping mental health sufferers set goals and improve their own self management. The Mayo Clinic (2015) states that those who suffer with mental health problem who keep track of their lives can more easily identify what triggers or improves symptoms, as well as find days easier to get through as the have a set agenda each day.

From my own point of view I did not just want to focus on creating a to-do application. As result I wanted Carpe DM to stand out from the other applications on the market, this is where the idea of taking advantage of a Conversational User Interface (CUI) came about. A CUI is any user interface that mimics chatting with a real human (Brownlee, 2016). There are two primary types of CUI, voice assistants such as Apple’s Siri, and chatbots such as Slack’s Slackbot. Carpe DM’s CUI would come in the form of a chatbot called Casey (the third most friendly name as voted on by over 30,000 users).

Casey is envisaged to be more than just a chatbot, I feel that the power of machine learning and artificial intelligence could be used to create a truly powerful companion for those suffering with mental health, a companion that would intelligently adapt and grow as the user adapts and grows to certain situations. Due to time constraints and lack of knowledge in the area of machine learning, all of the succeeding mentions are purely hypothetical.

Based on the fact that isolation (localised and social) is one of the leading contributors to mental health alongside Dr. Sandra Micken’s (2015) hypothesis that mental health sufferers (men in particular) do not want to, or do not know who to talk to. A very smart, AI such as Casey is a plausible tool which could be used to add benefits to people’s lives daily, as after all CUI’s mimic human conversation and verbal interactions, both of which have been found to improve mental health sufferers quality of life (Mental health in Ireland, 2007).

As well as tracking their day to day activities, emotions and habits, Carpe DM will be integrated with the ability for the user to track their mental health using a MHI test (discussed in the previous blog post. The MHI test is a quickly applicable, highly regarded test which could be administered through Casey. Tracking this data would be highly valuable to give a more accurate reading of the users mental health status over time and would be of particular interest to medical professionals.

The Process.

The design and development of Carpe DM encompassed all stages of the design process, with much focus being given to the user research prior to any major design effort.


Before design began, a total of 4 interviews were carried out in order to guage, measure and define the success of the project. 2 interviewees were mental health professionals (1 traditional counsilor, 1 mental health technologist) whilst the other 2 interviewees were sufferers of mental health issues. It was important to get a holistic view of both the needs of the end users of Carpe DM as well as getting expert advice on how to approach what is often seen as a taboo topic.

Interview Findings.

From talking to the mental health professionals it was clear that Carpe DM should be highly personalised to an individual user and follow ADAA topics, the application can be scored at the end of the project to determine project success. Another note of interest was the mention of daily tests to track ones mental health. This fits in with the concept of the MHI-5 tests.

Talking to the suffers of mental health help confirm my hypothesis that planning is a major mitigator for mental health triggers, with this information in hand it was clear that the ideas of trackers is something of benefit to the user.

Person Development.

Based on the findings of the interviews, two persons were developed as a means of empahy building and overall direction.


As with the personas in this project, scenarios were not the main focus, therefore could be developed in an ad-hoc way. This being said the scenarios created where to the best of my ability based on interviews and real life situations I observed. The scenarios created were highly persona driven. A persona is a tangible representation of the user that acts as a believable agent in the settings of a scenario (Cooper et al., 2015). Persona-based scenarios are concise narrative descriptions of one or more personas using a product to achieve specific goals.

Scenario 01 - Average Daily Use.

Sean has recently been diagnosed by a mental health professional with anxiety. He was recommended to use the Carpe DM mental health app during his time at college as to help relieve some of the stress and pressure he will inevitably be under. Early into his first year Sean begins to feel overwhelmed by everything in life.

Sean opens the Carpe DM application as recommended. In the app Sean quickly adds trackers to his day so he can arrange and plan his life accordingly. John adds a number of different trackers on a number of different days. John wants this task to be quick yet satisfying. When John adds trackers he expects alerts about them.

Scenario 02 - Finding Help.

Jane is a young professional working in marketing. Jane is under a lot of pressure in work to keep her job. Jane is a regular user of Carpe DM as her friend recommended it to keep track of her sleeping patterns as Jane fears she may be developing insomnia. Jane is at the point where she does not know where to go or who to turn to. She decides that she needs to find professional help.

Jane wants to be able to search and curate through mental health professionals as to find one that best suit her individual needs. When searching for a mental health professional she has a number of things in mind, price, background, accreditation and customer reviews.

Tone of Voice.

The tone of voice of an application runs throughout the entire product. Whether it be a news site or a betting agent, users expect consistency from start to finish. Tone of Voice guidelines can be as simple or as complicated as one wants, from simply outlining the style of language, to specifically identifying the punctuation to be used. Salesforce for example have a 12 page pdf covering the tone and voice they expect from their products.

With Carpe DM, I aimed to create a tone of voice guidelines that could be used to express the Carpe DM brand, set the application apart from competitors, build trust and influence people in the decision making process (Cummings, 2013). To do this I followed a simple template that was developed by Kevin Gilbert (2017), a content strategist at Domain 7. Gilberts template breaks down the daunting task of developing a tone of voice guideline into 4 simple steps; Your voice is, Write like this, Not like this, Why?. Below is an example of Carpe DM’s tone of voice.

Chatbot Flow.

The chatbot flow is the most complex step of this framework. Due to time restraints a full flow void of “dead ends” could not be completed, however below is an example whiteboard sketch of the CaseyBot sign up flow

Following the above four steps, allowed me to create a chatbot that could be interweaved into Carpe DM allowing users to perform tasks in a more engaging, more enjoyable, more meaningful way.

Iteration 01.
Iteration 02.

After the final designer were complete and Carpe DM was a functioning prototype in invision, it was then tested with 2 users as a traditional usability test with some follow up questions to gather qualitative results. It was also tested with a mental health professional to decide its real world uses based on the ADAA rating scale.

The Result.

As mentioned previously, the ADAA rating is a widely accepted measurement for the effectiveness of technology in mental health. My goal was to have Carpe DM score an overall rating of 3.5 on the ADAA scale, proving itself to be genuinely useful for those suffering with mental health problems. To administer a review and rating I asked a mental health professional who was familiar with working with the scale. The results are as follows.

Overall, the evaluation carried out by a mental health professional shows that Carpe DM has scored a total score of 4.2. Based on the original goal of 3.5, this shows that the application as a whole was a success in the eyes of professional opinion.

Application Walkthrough.

Thanks for
stopping by!

If you would like to find out more about my work or would like to work on a project together, then please reach out.

I am currently looking for interesting, rewarding roles in UI/UX and Product Design.