Happy App

Happiness and mental hygiene are intrinsically linked outcomes derived from ones own Neuropsychiatry and Social Cognitive Behavioral (SCB) development.  Through this lens, “learning behaviors”, along with their associated biological processes, should be considered, collectively, as a social determinant of health.  Unfortunately, there are a multitude of factors that can influence (1) how we learn, (2) to what extent we can convert new information to knowledge, (3) how well does new information align with previously learned knowledge,  (4) how does interpretation of knowledge influence our overall behaviors and (5) to what extent do those behaviors have an impact on global cognitive function, and (6) how does it positively  contribute to an individuals mental wellness.  My learning engineering solution can address the unique learning challenges of civic engagement through the use of artificial intelligence (AI); and by integrating a civic engagement curriculum into a Machine Learning  (ML) algorithm, it can provide a framework of policy recommendations by which AI-enable tools should operate, thus creating, an industry standard. 

The emergence of AI-driven Large Language Models (LLMs) can actually serve as a signal transduction mechanism for observing these “learning behaviors” on AI-enabled tools while serving as a large scale assessment tool to populate a model of healthy behaviors.  In turn, the resulting model from the data can serve as a framework for AI-driven systems and products.  Why is this important?  The more we interact with digital devices, the less we interact with information in real world situations such as in traditional classroom teaching or on-the-job training.  Increasingly, AI-driven digital tools serves as the medium by which we communicate with other people in order to learn and work.  Concerns arise when that digital medium does not accurately interpret or contextualize data in an accurate or conducive manner in which we can definitively rely on its accuracy and veracity for critical decision making.  If conveyance of behaviors, such as tone, can be lost via a simple written text message, what then are the implications when we apply AI technology to complex systems like medicine or jurisprudence?

Identifying patterns in “learning behaviors”, with the use of  ML, can provide a holistic understanding of SCB development, identify inequities in citizen development, assist with the creation of targeted solutions and provide opportunities for early interventions. The same underlying neurobiological mechanisms of the decision making process can be replicated by AI with an SCB taxonomy that can interpreting behavioral data.  Currently no such taxonomy exists, however, my solution introduces a theoretical classification system.  These behaviors ultimately drive our biological systems towards abstract ideas such as job satisfaction and self-actualization (the fulfillment of one’s potentiality).  Before self-fulfillment needs can be addressed, as defined in Maslow’s Hierarchy of Needs, a strong foundation of skills that address Psychological Needs must be built.  Given that many of these needs are abstract and complex in nature (ie. Social Needs, Intellectual needs), it is my hope that my solution can quantify these variables in order to achieve data driven personal development and success.

The Happy App is a personal development mobile application and tracking tool designed to create a human-centric, purpose-driven experience while learning and working across web applications.  Chatbot technology now provide a mechanism by which non-technical users can communicate with AI-enabled devices for the purpose of command execution or creativity.  This is the first time in the evolution of artificial intelligence where laypersons can have a direct and meaningful interaction with digital products. Leveraging the ease by which technology-based solutions can be now be created, I seek to utilize AI for the purpose of studying human behavior using a supervised and unsupervised learning platform designed to operate within a gamified environment.  The platform built upon this system is called “The Happy App”.  

Through the users interactions within the application, via activities and skill tests, the in-game character “Happi”, begins to change and grow.  Happi is actually an animated digital assistant designed to adapt to the user for the purpose of personal development, tailored specifically to the needs of the user.  The needs of the user are defined by the data collected via the API and begin to populate a psychographic profile of the user.  By providing direct insight about user behaviors, tracking and storing durable skills, mental stressor can be reduced by the application and allow more automations to be unlocked for the users.  The incentive to continue using the application is that this technology begins to organize and automate their real world processes and activities.  Over time, there is a slow reduction of distractions from overwhelming workload tasks and a reduction of excess environment stimuli in digital spaces. 

Traditional classroom learning for most K-12 students occurs in a linear direction with skill building and pre-determined benchmarks for testing knowledge.   Unlike traditional learning, the SCB skills required for civic engagement begin in early childhood (both in and outside of the classroom) and are carried into adulthood.  These skills actually develop in a manner similar to the way adult learning upskilling, in an agile workflow structure along with reinforcement learning and contextualization within their own knowledge base.  This system is designed to leverage the natural learning pathways we used in order to create our human intelligence.  By targeting learning behaviors, the AI’s machine learning algorithm will engage with the user utilizing a large language model (LLM) to create system prompts.  Unlike the popular 1990’s toy, the Tamagotchi, system prompt to care for the Happi character are, in fact, in-app notifications for tasks the user needs to address which have a direct, net positive effect on real world needs of the user.   The more the user and the character grow together, the more complex the AI-behaviors become.  In this way, the AI system, through the aggregation of deidentified data on the platform, the “intellectual capability” of the AI/ML model will become more reliable, thus more trustworthy when accomplishing tasks broadly across the user base.

The process of learning requires attention and focus.  While the internet is a cheap, readily available source of information, business practices across poorly regulated technology minimize the capability this resource can provide for nearly every type of learner.  Many, if not all, of the current social problems in the United States have a direct relationship to the data and information utilization from the web.  Therefore, in order to create more positive neurological behaviors, we must approach optimization of human intelligence by targeting the process of learning itself.  Happy App is designed to align AI-behaviors with those that do not cause cognizant dissidence in its users.  In fact, the same model used to taxonomically classify the behaviors of the users is the same process employed by the recommendation algorithm when assessing the appropriateness of the data being presented.  

As AI-enable tools and products become evermore present in our everyday lives, many ethical concerns become apparent, despite the complex understanding required to precisely articulate what is driving these concerns. Just as there are innumerable factors influencing SCB, there are innumerable factors that have the potential to detract from good mental health which has a direct link to cognitive potential and is an indication of future success.  The goal is not to eliminate all of these noxious elements on the internet, but identify barriers, provide insight, refine critical thinking and decision making skills with a clearer understanding of information presented in a personalized comprehensive manner.

This educational tool targets the data input to automatically generate personalized educational responses for the use of self-driven learning and personal development.  Through AI-assisted decision making, the creation and execution of actionable change can be enhanced, allowing users to gain core social-emotional learning skills and autonomy when choosing their best path forward on their educational journeys. The more complex the critical thinking task of the user becomes, more complex automations generated by AI become possible.  This shows respect for individual intelligence by supporting educational equity, within the bounds of individual intellectual capability.

The over goal of the Happy App is to align natural human behaviors with AI tools users need in order to facilitate more meaningful interaction with information they need.  By fostering a healthier and more cohesive internet environment, I hope to encourage self-directed learning through indexed paths of best-practices and reducing barriers that create digital information inequities.  My solution is direct-to-learner AI-driven tool that progressively adapts to the user, and integrates more features and automation options as the needs of the user changes. 

This solution allow AI to passively learn about the user, in order to provide a more personalized learning experience for the sole purpose of benefiting the user directly. Natural learning processes are targeted for the AI algorithm to understand the social cognitive behaviors of the user. This will pave the way for the model to be used as an early neuropsychiatric diagnostic tool and, hopefully, a non-pharmacological solution for Neuropsychiatric disorders.  

The Happy App was recently a semi-finalist in the DARPA AI-tools for adult learners competition. It was important for me to specifically seek funding through a government grant program because of the scale of benefit I am setting out to create. The competition was optimal for high-risk, technology based products with the goal of serving the greater public. For nearly two decades I have used these principles when teaching K-12 students, disable students, my surgical patients and physician colleagues. I have developed this system over a long time through my education in Neuroscience, Medicine, Clinical Research, Health Policy and Health Behavior. The submission process, for the better, forced me to critically assess each component of the system and how best to interact with each other before beginning to create a theoretical unifying model for NeuroPsychiatric and Social Cognitive Behavior. This, in itself, justified the feasibility of what can be achieved from the data processed in the model. This was not easy and is the precise next hurdle OpenAI (the developers of ChatGPT) are seeking to accomplish. From experience, it is quite rare to find physicians or behavioral experts that understand the manner in which our own intelligence is formed; from early Neuro-embryology up to, and including, understanding the SCBs which influence cognition and Neuropsychiatric pathology. My solution is not only timely, but it provides translational scientific applicability to artificial “intelligence” - a crucial perspective which is severely underrepresented in the software development community. Throughout that recent competition, we were gifted product development tools and business resources which allowed me to formally create my business. I am hoping involvement in these challenges will lead to opportunities to be invited into an incubator program. As a digital application, I have the freedom to relocate in order to continue design and development of this product. Currently I am the only person operating my business and it is time to scale up and orchestrate the resources needed to bring my concept to life. In addition to my clinical expertise, I have over three years of experience directly implementing health programs in medical practices and large hospital groups across the Northeast United States. I not only acquired senior managerial business and operational management training, I also designed and implemented a new operations infrastructure which was adopted across all regions leading to a efficient implementation process that reduced timelines through a streamlined workforce development process. This application is a generalized learning tool capable of adaptation to any curriculum. I am currently poised to take immediate action on rapidly developing my business, prototyping and deploying this solution. While my primary objective is to right-size AI-driven tools to operate in a net positive manner while improving mental health outcomes of its users, I very much look forward to seeing a bright future of equitable educational access for learners of all ages. If we can work to reduce barriers that create intelligent, civic-minded, productive workforce, the possibilities for the future of our fellow citizens is endless.

What specific financial, technical, legal, cultural, or market barriers that you face do you hope Solve can help you overcome, and how? Keep in mind: Solve aims to connect all Solver teams with partners who can help advance their solutions through monetary and/or non-monetary support. If your only goal in applying to Solve is to raise funds, Solve is not the right program for you! Review the Become a Solver page (“What will I get if my solution is selected?”) for more information about what selected Solver teams receive. I am apply to Solve for more than just an infusion of much need financial capital. As a newly formed LLC, and self-funded, I am in the very early stages of developing a fully-conceptualize mobile application and software system. Access to human and technical resources is my biggest priority. I understand the complex of the solution I am offering and have experience piloting, implementing, executing and evaluating programs in the health and wellness industry. I am a physician serving as the subject matter expertise across the entire academic spectrum of features I have designed for this product. My capabilities in the software development space are extremely limited, however, I have engaged with many IT teams throughout the Northeast US, for the purpose of training on Electronic Medical Records system and crafting personalized digital charts across multiple medical specialties; predominantly Surgical specialties, Emergency Medicine and Family Medicine. Creating methodologies to adapt to different clinicians, patient populations and community needs is a particularly unique talent that I have attained through engaging, training and supporting diverse learners in both clinical and non-clinical groups. Simply put, I think that participating in the Solve program would be particularly beneficial for a founder and product as this stage of development. Over the last two years I have taken over 50 online courses to fill the gap in my Computer Science knowledge. I understand intelligence systems in biological systems which as made filling my lack of a formal academic education in artificial intelligence easier to comprehend. I am one person and need to create a board, and jump start my business. I would benefit immensely from the precise resources collated in an incubation program. I am at an inflection point where my theory has been fleshed out and is primed for production. I am in need of every resource from a software development team, data analysts/scientists, a research coordinator along with a base of operations and production tools. Lastly, I will need legal assistance for Intellectual Property (IP) protection and business management. I have completed the training modules for inventors through the US Patent Office (USPTO) and plan to craft and file utility patents for the system design. I am currently on the waiting list for the Bar Associations National Clearing House Pro Bono program and have selected two regional patent offices based on my location and regional office specialization. I would like to expedite this process as it has been nearly a year (at the time of this submission).

My solution utilizes a large language model as a mechanism to interact with a unique ML algorithm designed to understand learning behaviors through the use of a novel taxonomical classification system. Treating online behaviors as a social determinant of health and as data that should be protected under HIPAA policies. Only to be utilized for the direct benefit of the user. Unlike traditional digital assistants, this one is specifically designed to adapt to the learning behaviors of users in order to maximize information comprehension based on individuals personal intellectual needs. The gaming features and the Happi avatar provide visualizations of skills and achievements to support mental wellness and contribute to higher educational planning and/or career exploration. My solution is a direct-to-learner tool used outside of the classroom on mobile devices that allows students to engage in self-driven learning across web data as they interact with search results. Providing users with personal insight on their behaviors as an early intervention tool to promote self-awareness and creating opportunities to create better behavioral patterns supported by AI-driven tools. Social tools that only become active through maturity of the user on the platform, addressing a fundamental flaw in current social platform designs. Incentivizing users to make real world changes for themselves and their community.

Goals for the next year include establishment of a reliable source of financial funding (critical in the next 3 months) during the development phase. Ideally I would like my business to be developed in an incubator program to provide people and technical resources to prototype as it will likely be a minimum of 12 months before there is a fully marketable product. Beyond year one I’ve created a plan to use the prototype, pitch deck and web landing page as resources to seek AI or VC capital funding. I plan to establish a public/private partnership in which education and government services are the first data sources to populate the model and provide stability for the behaviors of the application. Once vigorous testing is complete, I’d like to have pilot studies in place by year two. Ideally beginning with middle-school aged children, then progressing to high school. By year 3 I would like to have educational partnerships, fully deployed applications spread over multiple regions. Social contribution features will be established, I have a novel solution to address this. A SaaS model should be produced in order to pilot curriculum creation. Work force development tools will also be integrated into the platform. By year 4 I plan to introduce higher educational models and by year 5 there should be broad public applicability with established users across multiple markets with some paid features and a fully operational SaaS model with low-cost subscription plans for maintenance of the courses. If this proves successful, I have a broader business model to scale into operating system development.

The Transtheoretical Model (TTM), or Stages of Change Theory, developed in the 1970’s describes the comprehensive process of behavior change. It is utilized in modern Psychotherapy and Preventative Medicine to improve health behaviors, however, it has equal applicability when generalized for any learning scenario. Behaviors, like information, can be learned, improved upon or built around to reinforce. The process follows a predictable pattern of stages: (1) Pre-contemplation, (2) Contemplation, (3) Preparation, (4) Action, (5) Management and (6) Termination. This simple stratification can help index resources, information, webpages and/or actionable items along easy to follow paths where the end-goal is having acquired the new knowledge or skill. Complex subjects are typically broken down into digestible subunits (lessons) and taught in a linear direction. Typically, pre-contemplation serves as the traditional entrance point for most new learners via some exposure method (ie. Advertising, Notifications, Instruction, etc). Rather than a linear direction, adults often learn in loosely structured formats with frequent branching, mimicking an agile workflow. Skills are often learned in tandem or overlap disjointed timelines. This is due, in part, to adult learners typically having some previous knowledge related to the subject matter. Therefore, learning can also begin in later stages of the TTM model. Adapting to this model, civic learning must be engineered to act like a game of Dominoes, slowly building on established knowledge, assembling a chain. New skills completed in one course can then be connected to the subsequent lesson or incorporated within a completely different curriculum. Pre-contemplation also describes the actions performed to get the user to engage in the very first task of a new learning trail. For subjects taught in isolation, similar to a jigsaw puzzle, knowledge and skill building can begin anywhere. The process of “funneling”, whereby new skills and knowledge are built upon through traditional learning, progresses linearly and into the next subject. Completed skills can reverse funnel, connecting back to earlier lessons. By designing various engagement methods to enter later stages of TTM, learners are more likely to adopt the related skill than when taught alone. A Software as a Service (SaaS) model can offer opportunities to create “Learning Levers”, similar to “stop” prompts, in order to facilitate smoother transitions between related topics and test the knowledge of the user before the concept is introduced, creating critical contextualization while gather additional user data; all to propelled self-directed learning. Alternatively, learning paths can be imported from a searchable indexed library to spark interests in other subject matters. By using ML, current skills can be connected to the new skill by plotting a path across multiple learning trails to populated individualized courses for users. Like ML-guided image recognition, learned skills and insights generated by the AI can be used to extrapolate a detailed picture of the user's learning needs. Websites currently utilize trackers and cookies in a similar function to personalize the web browsing experience, however, this method seeks to create a highly bespoke experience while offering a greater level of privacy protection.

AI-driven tool, Gamified AP, Machine Learning Algorithm using a theoretical model for taxonomical classification of learning behaviors. 

Web browser

- Web crawling tools

- New navigational options to view search results with personalized recommendations 

Security - end-to-end encryption

- Block chain security

- HIPAA compliment (Neuropsychosocial behavioral data is treated as medical data)

As the app develops and use of the platform increases, the user will have opportunities to personalize their Happi avatar with ethnocultural characterization allowing users to integrate lived experiences into the platform.  Through self-identify of these diversity related characteristics, users are actually identifying additional variables by which we can better analyze inequities across the datasets.  The Happy App is intended to gradually engage with learners in order to combat barriers to information typically provided through tradition web services.  With the use of a web-crawler LLM feature, sites can be tagged for appropriateness to the user, have the accuracy of the information assessed which can then lead to a new generation of personalized recommendation algorithms.

The goal of the Happy app is to understand behaviors in a meaningful, quantifiable manner to which it can then aid in early recognition of cognizant dissidence between the user and the SCB database being generate by the model.  Utilizing insight gleaned from other data, imported from other social platform and mobile applications, it is theoretically possible for AI to no only recognize less pervasive forms of cognizant dissidence, but even anticipate poorer outcomes and recommend solutions to curb behaviors before they become pathological.  Through the use of a continuous data monitoring, the Happy App can provide support for users in near real time, decrease mental distractions, assist in the decision making process all while providing a means for the user to achieve personal development.  By identifying unmet needs, we can improve the various components of our own intelligence over time.  By this mechanism, the platform strives to be inclusive, as best as possible, of users with pervasive developmental disorders and NeuroPsychiatric conditions, with a focus on those that relate to attention, language utilization and cognition.  More research will be needed to design a broader implementation plan alongside the product development in order to achieve these goals before the public launch of the product.

My business, Happy Laboratories, LLC was recently incorporated as an C-Corp, LLO.  Currently under SAM.gov review to receive reward money should I win.  This is a for profit business model but as we begin to population the model data, we will create a public/private partnership where the government my used the services at reduce cost

my complete business plan was created when I incorporated my business. More details can be submitted at a later time.

I have been self funding while submitting my first Challenge submission for DARPA’s AI tools for adult learning.  As a semi finalist we relieve additional resources for organizational development