EcoChum

AI-powered green platform: Conquer climate-related mental health challenges. Find joy in conversations, make an impact with personalized climate initiatives. Elevate your well-being and the planet!

Comment

Inspiration

EcoChum was born out of my personal journey, as I, Vinamra Sharma, ventured from India to Scotland to pursue my masters in computing science. The stark contrast in climate and unfamiliarity of this challenging new place left me isolated, with no friends or family. Days faded into darkness by 2pm, accompanied by perpetual rain. It was then I discovered the profound impact of climate on mental well-being. This emotional experience ignited me to make sure that no one else should have to face the same. The Climate Change Maker Challenge presented a defining moment to craft a purposeful solution, where we met our teammate Ayumi. With an unwavering drive to address climate issues, Ayumi seamlessly blended design and a human-centered approach, shaping our initial structure with passion and purpose. Heba was one who have personally evidenced the issue of climate change while living in Canada and seeing how all of these situations were impacting the lives of individuals. Equipped with the understanding of sustainability and business baked with computing background extended her interest to define the orientation for EcoChum. Whereas our team was incomplete until Ariana added up and extend the technical and design infrastructure with the sense of peer compassion coupled with deep physiology and human interaction.

What it does

EcoChum is an integrative green platform helping people combat mental health struggles related to climate-change by providing you a happy chum(ChatBot) to talk to which suggests solutions that uplift's individual's wellbeing by suggesting involvement in climate change initiatives that best suits the person.

EcoChum offers a wide range of services and features that include the following:

  1. ChumChat: The AI chatbot(Large language model) suggesting eco-friendly ways to deal with the issues and promoting community involvement.
  2. GiftChum: Allows you to gift your friends and family a tree planted specially with their name; this will be a token enriching their good moment and nature at once.
  3. EventChum: One stop place to find all the green/climate-focused events going across the place, so that you do not miss out on things and get involved with individuals driven with a meaning of bringing a positive change.
  4. EcoVoices: A library of motivating series/videos/voices that fills you with the right motivation required to bring the positive chnage you always wanted to.
  5. EcoBlogs: Allow you to explore what's going across the world of eco-warriors

An interesting part about EcoChum is that; it is alive. Just like you and me, it is learning from each and every user interaction. On an attentional part we considered security as our key component in the development (this is generally neglected when it comes to climatic data); All the data is strictly encrypted using the PGP protocols and the servers are isolated consuming the firewall and external routing.

How we built it

Besides the technology we have used, the EcoChum is made with our emotions, determination and sense of brining the change we always wanted to. In order to define the development of the model, the following three aspects are considered:

  1. Technical Built
  2. Business Prospect Built
  3. Human Extensive Prospect Built/ Research Built

Being specific in the technical terms; we have consume way lot of python to define the Language model and have employed a wide range of libraries to get the task done. The web platform consumes WordPress, React.Js and Node.Js. While external factors includes Google Cloud for the Dialogflow integration and deployment of the chatbot, the website is deployed over Linux based servers that are capable of multi-node threading. The OpenAI API's are used to define the initial intents, while later Dialogflow ML intent training is consumed to further extend the interaction.

In term of the business prospect; we have defined the user orientation, target audience, market strategy, potential stake holders, and organisational tie-ups. Though the initial plan was very general but connecting with mentors helped us to gain the clear understanding of the loop-holes in our business prospective. Monika helped us understand how we can make the platform effective by adding up potential numbers to it which we delivered by add the Climatica Impact Board section in the dashboard.Monika suggested a lot other great points that we haven't though at initial stage. Later, Pam helped us a lot by mentoring regarding the approach to consider and land to right stakeholders as well as how we can incorporate them well to the project and even beyond how we can take this further in the real world(which we will definitely aims to do), she also extended her help by sharing more accurate resources and research pointers to deeply examine the market and implementation landscape as well as how to go across the funding.

The the final part; we have did both quantitive and qualitative research to check the problem landscape and validation. We have used forms to collect the user input on the problem and we have tried to incorporate the features that most of the people who have went through the issue employed to overcome. In this part we are very much helped by Shitangshu and the whole organising team with the information shared through the 1:1 sessions, workshops, and even the motivating fun events. The validation of solution is considered based on the feedback received from the mentors as well as a lot other factors are discussed and further extended with external mentorship by Garisha.

Challenges we ran into

Well, the whole project journey was fun and enriched with learnings; Few of the challenges that we ran into are as follows:

  1. Managing the project effectively over different time zone and International team based over different countries.
  2. Implementing the technical stack; as the project is quite big and the time was limited the final implantation was a great challenge. Working onto the language model was highly resource intensive.
  3. Incorporating the tech, business and user personas together and define a solution that is a good trade-off for all three.
  4. Integrating the psychological frameworks in the algorithm development and user/market study.
  5. Finally compiling everything together while everyone is in middle of final exams.

Despite of these challenges, this journey was great as the whole team had an amazing understanding and not even a single meeting we had felt like a meeting but was more of a destress from our day to day life.

Accomplishments that we're proud of

  1. We are actually proud of everything that we did, and more of that we are able to compile all our imagination into a working prototype.
  2. We are happy that we have connected as a team and in their exciting journey meet with great mentors, fellow hackers, amazing organising team and made a network not just for event but for life.
  3. We have exchanged our skill sets and learned from each other.

What we learned

We have learned the skills of collaborative working, management, project evaluation, research, market prototyping, development, complex implantation, user centred design, and combining resources effective as well as efficiently. We have learned that when you have passion and determination with right guidance and people; all the challenges fades away and the light of meaning outshines.

What's next for EcoChum

Well we actually have a lot of plans for EcoChum including:

  1. Further extending the language model to generate more defined/accurate answers.
  2. Collaborating with the EarthNet team (possibly) and mutually enriching the work.
  3. Further extend our team and work to take this to the market and evolve into a sustainable ecosystem.
  4. Extending the technical aspects and launching the beta version for the world.
  5. Evolving further based on the user interaction and feedbacks.
  6. Looking for potential funding and exposure to take this to next step.
Share this project:

Updates