Scholé: Upskilling for Adult Learners

The rapid advancement of technologies under the umbrella of Industry 4.0 has created a seismic shift in job requirements and the skills necessary to fulfill them. We directly aim to address the gap presented by the 2023 US National Defense Science & Technology Strategy report stating that "we cannot create 21st-century capabilities using 20th-century education practices", DARPA's "Building an Adaptive and Competitive Workforce" challenge leveraging LLMs and GenAI towards adult learner needs in STEM and data science, and more broadly, the World Economic Forum (WEF) prediction that by 2025, "50% of all workers worldwide will need reskilling to compete in the market".

The engagement in lifelong learning is notably low. For example, a European Commission Report states that less than 40% of adults participate in ongoing education annually—a rate insufficient to meet the demands of Industry 4.0. The popularity of distance learning approaches has been accompanied with a crisis in student dropout, where MOOC completion rates are often less than 13%. Such educational deficits pose a critical barrier to personal and regional economic development, particularly as high-tech industries continue to expand their influence over the global economy.

For communities specifically targeted by Scholé, the challenges are manifold and include:

• Mid-Career Skill Gaps: Many adults in the workforce are finding themselves ill-equipped for the contemporary skills demanded by their professions, especially in fields rapidly integrating new data science and AI technologies.

• Language Barriers: The prevalence of English as the primary language of technology and business poses a significant challenge for non-native speakers, limiting their access to necessary educational resources.

• Cultural and Educational Diversity: Educational resources often fail to be inclusive of the diverse backgrounds present within many urban and immigrant communities, leading to a lack of engagement and educational attainment.

• Flexible Educational Content: Most available educational content require a larger level of time commitment over weeks or months. Adult learners often have tight schedules with varying amounts of time (20 minutes before a meeting), or varying modality availability (audio-only podcast for a commute). 

Several key factors contribute to the urgency and complexity of the reskilling imperative:

• Technological Advancement: The pace of technological innovation means that the “half-life” of professional skills is decreasing. Workers must continually acquire new skills to remain relevant in their fields.

• Economic Shifts: As economies transition from traditional manufacturing to high-tech and service-oriented industries, workers without the requisite tech-savvy are increasingly at risk of displacement.

• Educational Inertia: The existing educational systems are often slow to adapt to the changing landscape, creating a lag between emerging industry requirements and available training programs. There is a growing population of learners that want the flexibility that conversing with a model provides, while also receiving instruction and generalized learning outcomes.

The gap between current workforce skills and those needed to achieve economic and professional success in the future is widening. Scholé seeks to mitigate these challenges through adaptive, personalized learning technologies that respond to the unique needs of adult learners across diverse linguistic and professional backgrounds.

In the context of adult education, where time is a crucial commodity, it's essential that the learning experience is both efficient and directly applicable to the learner's professional context. Scholé caters to the upskilling demands of various professions by focusing on the practical application of data science skills tailored to the specific requirements and preferences of each learner. For example, analysis of clinical trial data is more relevant for healthcare professionals whereas demand forecasting is better suited for professions in the supply chain sector. 

Scholé stands out by enabling scalable personalization for each learner including customized career-centric curriculum and problem sets through finetuned LLMs. Using these models, we also enable data science education in different languages (e.g. Spanish), reaching learners of all backgrounds. Scholé offers a tailored learning experience through two distinct tracks: the Goal-Oriented Track and the Profession-Oriented Track, designed to cater to immediate learning needs and broader professional upskilling, respectively. Progress in either track is synchronized, ensuring that skills learned in one area benefit the other.

The Goal-Oriented Track is for learners with a specific problem or goal in mind. Users can input their problem via text (e.g., copy-paste), voice, or screenshot, and an AI-tutor then crafts a personalized learning path, emphasizing the necessary skills and knowledge to solve the problem. Learners can adjust the depth of their study in each module, with the option to take quizzes to shorten the learning path.

The Profession-Oriented Track is aimed at users seeking to enhance their general data science skills applicable to their professional field. Through interaction with the AI-tutor, learners outline their profession and responsibilities, and a custom curriculum is developed. This curriculum visualizes a learning path with interconnected modules, each detailing learning outcomes and dependencies. Learners have the flexibility to rearrange the path and adjust the depth of modules based on their interests, with each module's "length" adaptable to show the level of interest.

Both tracks offer a co-designed curriculum with the AI-tutor. Then, based on extensive SRL research, the AI-tutor will lead the learner to plan their studying time and set timeline goals to finish each module. A "Competencies" shelf tracks progress in specific areas. Learning materials are interactive, with quizzes, videos, and coding assignments tailored to the learner's context and preferred programming language. The lesson material, as in other ITS, adapts to the learner's level, offering challenges appropriate to their skill level.

Scholé recommends personalized curriculum based on a directed knowledge graph where nodes represent data science topics and edges denote prerequisite dependencies. Constructed collaboratively by a team and advisory board utilizing extensive online educational resources and textbooks, the graph is also updated based on user feedback, market trends, and technological advancements, allowing for the integration of new modules. LLMs analyze user inputs (text, voice, screenshots) to communicate with the recommender system. The learners' skill levels are assessed through knowledge tracing models (e.g., graph-based knowledge tracing) while Generative AI tailors materials and quizzes to the learner's context and skill level using scaffolding and Bloom's taxonomy. 

Scholé is designed to serve adult learners in the workforce who are facing the challenges of adapting to the rapid technological shifts of Industry 4.0. This target population includes mid-career professionals who are increasingly finding that their skills do not meet the new demands of their fields, particularly in sectors integrating advanced technologies like data science and artificial intelligence. These learners are typically underserved by traditional educational systems that are slow to adapt to the fast-evolving job market and its requisite skill sets.

Therefore, our solution focuses on several key demographics that are particularly vulnerable:

• Mid-Career Professionals: Individuals who have been in the workforce for years and need to reskill to avoid displacement due to the shift towards high-tech and service-oriented industries. Scholé offers them tailored learning paths that update their skills relevant to their current industries and job roles.

• Non-native English Speakers: With English being the dominant language of technology and business, non-native speakers face significant barriers in accessing educational resources. EduContext addresses this by providing multilingual support (including English), ensuring these learners can access high-quality data science education in their own languages.

• Culturally Diverse Immigrant Communities: EduContext's personalized learning environments adapt to diverse cultural backgrounds, making learning more inclusive and effective.

Scholé's impact is structured around providing personalized, flexible learning experiences that align with the learners' specific career goals while accommodating their unique schedules and language needs. This is achieved through:

• Goal-Oriented and Profession-Oriented Tracks: These tracks allow users to either focus on immediate, specific learning goals or undertake broader professional development. This structure supports learners whether they are looking to make a lateral move within their industry or advance to more technical roles.

• Personalized Learning Technologies: Using AI-driven recommendations, the platform customizes the curriculum based on the user’s professional background, proficiency, and learning pace, thereby enhancing the relevance and efficiency of the education provided.

• Community and Stakeholder Engagement: By partnering with employers, community organizations, and educational institutions, Scholé ensures that its curriculum remains relevant to current and future job markets. This engagement also helps to refine and adapt the learning experience based on comprehensive feedback from a diverse user base.

Overall, Scholé aims to provide the necessary tools for adult learners to adapt and thrive in a rapidly changing job market and ensure that these opportunities are equitable and accessible to all, regardless of their socioeconomic or cultural background. By doing so, it transforms individual career trajectories and contributes to broader economic development and social inclusivity. 

The core team members are Vinitra and Paola, technical co-founders and the first two graduating PhDs from the ML for Education Laboratory at EPFL. Both team members will go full-time on this project post-graduation.

Paola's PhD is on self-regulated learning (SRL) in online or blended environments. She has worked closely with vocational education system in Switzerland, developing a strong connection with the Swiss State Secretariat for Education Research and Innovation (SERI), in particular the Division for Vocational and Continuing Education, attending their last two conferences. From a technical perspective, Paola has worked with ITS in start-ups (Lernnavi and Skillpartout) implementing new features to engage users (e.g., SRL adaptive messages based on student profiles, team features) and promoting SRL in mathematics and language learning. She has multiple projects with vocational schools across multiple professions (professional education), including the creation of a digital tool to enhance writing skills for chef apprentices and reflective journaling for nurses. Her experience working directly with learners and teachers at trade and vocational schools, especially in the context of upskilling with AI for education solutions, is extremely relevant to desired user base of EduContext.

Vinitra worked on scaling UC Berkeley's first data science course from 100 students to 1700 students a semester with students from over 60 major disciplines, 10 context-adapted connector seminars and a 50-50 gender ratio. She then lectured the flagship data science offering at UC Berkeley in summer 2018, as well as an intro ML course at UW in 2020, showing her familiarity with teaching data science concepts at the higher education level. Her PhD is about explainable interventions for education, providing personalized and tailored assistance to students in online courses, especially in contexts where teachers are not present. She has conducted several validation studies with professors teaching MOOCs (massive open online courses) and worked with clickstreams of distance-learners for nearly all of her projects in the PhD, showing immense familiarity with the student data that is relevant in building a solution like Scholé.

Both Paola and Vinitra have been head teaching assistants for three iterations of the ML for Behavioral Data course, co-designing the curriculum, projects, and teaching materials. Vinitra's familiarity with the American educational system and Paola's familiarity with the Mexican educational system, in addition to their teaching experience in Switzerland gives them a global perspective on education that they will use towards Scholé, as the tool targets international B2C personalization. Most notably, their connections to different groups of vocational student populations in Switzerland (chefs, pharmacy assistants, medical trainees) will be apt to deploy formal pilots of the tool to measure learning gains.

Scholé is closely partnering with Alejandra Domenzain at UC Berkeley, a Program Coordinator with the Labor Occupational Health Program (LOHP) working on skill-building training for immigrant low wage workers (including domestic workers, day laborers, farmworkers, janitors, and restaurant workers) to reach the communities that Scholé would most help.

We have a minimal set of Scholé testers (5) that have participated in a series of design and usability interviews using our initial features. These interviews help us understand the best ways to present information and meet our users' needs effectively. Our efforts are focused on refining our prototype based on the feedback gathered from these sessions, ensuring that our solution integrates the users directly in the design process and is tailored to offer a meaningful and efficient experience.

Towards the development of the tool, we have components working individually (including an initial customized LLM towards personalized learning, a minimal knowledge graph of data science materials, and an explainable knowledge graph traversal algorithm). We would like to use the grant to improve these components and validate our knowledge graph of data science learning materials alongside our expert network of data science teaching professors. Additionally, we would like to further integrate these three components together (LLM chatbot, curriculum curation, knowledge graph of data science concepts) with a nicer UI alongside gamification and community features towards a fully functioning platform and public launch.

By partnering with Solve, Scholé expects to extend its reach and impact so that more adult learners have the skills they need to succeed in the rapidly evolving workplace.

Firstly, market access and industry validation are critical for the adoption of Scholé within professional sectors. Through Solve, we aim to connect with industry leaders and potential clients who can provide market feedback and pilot testing opportunities. These interactions could validate our solution but integrate the requirements and feedback from potential users. 

Moreover, the legal and cultural barriers in providing multilingual education to diverse demographics pose a challenge. Solve's network can provide crucial insights and partnerships necessary to navigate the regulatory landscapes of education across different regions. This is important as EduContext has the goal of serving native and non-native English speakers and integrates cultural relevancy into its curriculum, enhancing the learning experience for a global audience.

Additionally, scaling a platform like Scholé requires robust operational and financial strategies, beyond mere fundraising. Solve's mentorship can guide us in establishing a sustainable business model that supports expansion while keeping the platform accessible, especially to underserved communities. Insights into effective pricing models, cost management, and user engagement strategies from Solve's resources would be invaluable. 

Lastly, Scholé aims to overcome the technical challenge of personalization at scale with guidance from Solve. Scholé utilizes finetuned LLMs to create personalized curricula and problem sets tailored to the individual professional context of learners.  While we have both the technical skills and working versions of initial prototypes for our technology, collaborating with Solve can amplify our access and funding for SoTA language models and training resources. This partnership could particularly help us navigate the complexities of deploying AI in educational settings while maintaining ethical standards and learner privacy.

Scholé is an innovative learning tool addressing the need for personalized upskilling for adult learners, tailored to their specific professional contexts and learning needs. 

Our platform is distinguished by its ability to adapt learning experiences based on individual career needs and preferences, significantly enhancing the efficiency and relevance of educational content, based on SoTA AI in learning science research. We are initially targeting adult learners upskilling data science skills, but provide a broad solution that can be expanded to many different domains and learner groups. Scholé integrates two distinct learning tracks: the Goal-Oriented Track and the Profession-Oriented Track, both of which are synchronized to provide comprehensive skills applicable to real-world demands.

Key Innovations and Market Impacts:

1. Personalization through Generative Models: Utilizing LLMs and LMMs, EduContext adapts existing quality learning content to user's specific professional environment. This approach increases engagement by aligning content with learners' immediate needs and promotes that the skills acquired are directly applicable and transferable to their respective careers. It combines the learning benefits of existing, open source, quality learning materials with increased user engagement towards topics that interest them.

2. Dual Learning Tracks: The platform's dual track system allows learners to either focus on immediate practical problems or undertake a broader exploration of data science as it applies to their career field. This flexibility supports a wide range of learning objectives and adapts to different levels of learner commitment and time availability. Specifically, we target learners who know they could answer their question directly with a service like ChatGPT but seek to really learn how to solve their query in addition receiving a direct answer.

3. Multilingual Education: By providing data science education in various languages, Scholé extends its reach beyond English-speaking populations, thereby democratizing access to cutting-edge skills in high demand across global job markets.

4. Social Learning and Community Building: Scholé incorporates social features that encourage peer-to-peer interaction and collaborative learning. This community aspect fosters engagement and motivation among learners by enabling them to share achievements and receive support from peers with complementary skills.

5. User Adaptation of Customized Curricula: Interactions with our LLM chatbot EduCo allow users to specify exactly how they would like to change the proposed lesson plan: "more problems", "audio-only", "make it 20 minutes", "I don't want to do this part of the lesson", "change the focus to topic X". This enables great flexibility and a learner-centric approach in a world where online course dropout is extremely high and maintaining learner interest is key. We believe users, and especially adult learners, should be able to learn what they need, when they want to learn it, how they want to learn it.

Our larger impact goals are towards a vision in upskilling for underserved populations, creating equitable opportunities through personalization for a large global audience. The internet is a grand equalizer for information availability; we believe AI and ML for education can enable an analogous equality in learning quality.

The impact Scholé seeks to achieve is twofold: to increase access to career pathways for adult learners by providing them with the necessary data science skills, and to close the skills gap that limits economic growth. The theory of change is based on the premise that personalized, efficient, and directly applicable learning experiences can significantly enhance the workforce readiness of adult learners.

Scholé addresses the challenges faced by adult learners in the workforce by offering efficient, flexible, and relevant upskilling in data science. These challenges include limited learning time, the need for personalized learning experiences aligned with specific career goals, and overcoming language barriers in education.

A specific use case demonstrating Scholé's response to workforce barriers is its support for a mid-career professional seeking to transition into a data science role within their current industry. For example, a marketing manager in retail might engage with the Profession-Oriented Track to enhance skills in data analysis and consumer behavior analytics, while a paralegal could use the Goal-Oriented Track to develop document retrieval skills using LLMs and RAG.

Scholé engages with various stakeholders throughout the product design and development process through:

• Stakeholder Collaboration: Partnering with employers, industry experts, and educational institutions to align the curriculum with current and future job market demands.

• User-Centric Design: Conducting interviews with a diverse user base. Utilizing our partnership with Alejandra Domenzain at UC Berkeley, who works directly with immigrant, low-wage workers in a variety of high-hazard industries, we can ensure we reach the populations of users who will gain from upskilling the most. 

• Pilot Programs: Implementing pilot studies with target demographics such as non-native English speakers to refine the platform based on real-world feedback.

• Online Training and Human-in-the-Loop Models: Collecting and analyzing user feedback regularly to adjust features and improve curriculum relevance and AI tutor effectiveness.

• Accessibility Focus Groups: Conducting focus groups with learners who have diverse needs and disabilities.

EduContext adaptivity ensures that learners from diverse backgrounds can find content and problem sets that are directly relevant to their career paths, thereby increasing the applicability and impact of their learning. Furthermore, by offering multilingual support, Scholé breaks down language barriers, making data science education accessible to a wider audience. To maximize its reach and impact among underserved communities, Scholé employs several strategies:

• Affordable Pricing: Keeping the platform affordable or free (with scholarships) for users from underserved communities.

• Mobile Accessibility: Ensuring that Scholé is accessible on mobile devices, addressing the digital divide.

• Community Partnerships: Collaborating with community organizations, NGOs, and educational institutions within underserved areas to tailor the platform's offerings to the needs of these communities and to facilitate access to Scholé's resources.

Moreover, Scholé involves learners and stakeholders from diverse backgrounds in the creation process. By engaging with these communities through feedback sessions and pilot programs (as mentioned before), EduContext ensures that its development is guided by the understanding of the challenges and barriers these learners face.

We believe Scholé's approach could catalyze significant positive changes in the landscape of adult education by demonstrating the effectiveness of personalized, flexible learning environments. It highlights the potential for AI and machine learning technologies to significantly enhance learning efficiency, which could inspire similar innovations across other educational fields.

Moreover, by addressing language barriers and making advanced skills (especially starting with data science) accessible to a broader audience, Scholé has the potential to close the global skills gap. This could lead to a more diverse and capable workforce, equipped to tackle complex problems in various industries, thereby driving economic growth and innovation worldwide.

Scholé's model of using generative AI to modify learning materials towards relevance in adult education could inspire other educational platforms. As this model proves its efficacy, it may encourage a shift towards more learner-centric technologies in education technology markets. This shift could increase the competitiveness among educational platforms, pushing for further innovations in personalization and relavence in learning technologies.

The proposed metrics to measure the impact of EduContext include: completion rates of learning tracks, improvement in data science skill assessments before and after course completion, ultimately job retention and advancement rates among learners, feedback from employers on the performance and skill levels of employees who have used EduContext. 

We use three main models in Scholé: a LLM chatbot EduCo, a curriculum curation model, and a multimodal personalizer. First, we tailor a chatbot to determine the needs of the learner: what their learning goals are, how much time they have for the lesson, what kinds of modalities they prefer. We will use in-context prompting, instruction tuning, or ideally finetuning to refine a chatbot for Scholé.

Once we have identified the user's learning needs, we create a personalized lesson plan by identifying a relevant path over the expert-developed knowledge graph. Our curriculum curation model traverses our knowledge graph of data science concepts with a graph path explainable recommender system (UPGPR) along the lines of SoTA research (Frej et al., AIED 2024). This graph is grounded in a collaborative process that leverages expertise from educators, industry professionals, and fellow learners. This ensures that the learning content is academically rigorous and industry-relevant. The curriculum is continuously updated based on feedback, emerging trends, and new research, maintaining its relevance and effectiveness. We will audit both the knowledge graph and the curriculum curation model to identify the potential biases and mitigate them. 

Once the lesson plan is generated, we use our multimodal personalizer (e.g. Swamy et al., NeurIPS 2023) to evaluate the personalized curriculum for logical coherence and generate the adapted materials with the context of the user's career (i.e. changing a generic A/B testing problem into a relevant one for the e-commerce sector). The first model (chatbot) and third model (personalizer) will work together to modify the lesson plan to the user's feedback.

Scholé's effectiveness is validated through controlled lab studies and field experiments (Mejia et al. AIED 2023, Swamy et al., LAK 2023), incorporating real user feedback to enhance learning outcomes, engagement, and satisfaction. The platform is designed to be intuitive and mobile-responsive, accommodating learners of all backgrounds, including non-native English speakers with multilingual access.

The platform upholds high data protection standards and provides transparency about data usage. Explanations of the recommender system’s predictions are made available through XAI, ensuring ethical use of data and fostering trust among users. 

EduContext tracks clickstream and interaction data including changes to the learning path, module completion rates, knowledge tracing parameters, learner's progress, and interaction between peers. This data is relevant for the curriculum generation model therefore the "secret sauce" towards highly effective learning outcomes. Additionally, the user-specified problems (from the Goal-Oriented Path) will enable deep understanding of which data science needs are most relevant for which domains. 

The data science topics graph is continuously updated based on learner's feedback (modifying their proposed trajectories) and new materials. By leveraging learning analytics and educational data mining methods, we can identify behavioral patterns and profiles, predict learning challenges (e.g. how many more attempts a student will take on this problem, which skills need most help), and adapt the educational content accordingly. Scholé provides tailored interventions such as metacognitive SRL prompts to help learners stay on track and be aware of their learning.

[1]  J. Frej, N. Shah, M. Knezevic, T. Nazaretsky, and T. Ka ̈ser, “Finding paths for explainable MOOC recommendation: A learner perspective,” in Proceedings of the 14th Learning Analytics and Knowl- edge Conference, pp. 426–437, 2024.

[2]  V. Swamy, M. Satayeva, J. Frej, T. Bossy, T. Vogels, M. Jaggi, T. K ̈aser, and M.-A. Hartley, “MultiMoDN—multimodal, multi-task, interpretable modular networks,” Advances in Neural Information Processing Systems, vol. 36, 2024.

[3]  P. Mejia-Domenzain, E. Laini, S. P. Neshaei, T. Wambsganss, and T. Ka ̈ser, “Visualizing self-regulated learner profiles in dashboards: Design insights from teachers,” in International Conference on Artificial Intelligence in Education, pp. 619–624, Springer, 2023.

[4]  V. Swamy, S. Du, M. Marras, and T. Kaser, “Trusting the explainers: teacher validation of explainable artificial intelligence for course design,” in LAK23: 13th International Learning Analytics and Knowledge Conference, pp. 345–356, 2023.

Scholé Core team: 2 (Vinitra and Paola)

Educational Data Mining research collaborators: 3

Data Science Teaching experts: 3

Ethics board: 2

Advisory board: 3

Less than 1 year

Diversity is a core value at Scholé, which we enable by integrating a wide population of users at every step of the tool's design process. 

Our core team is two female, technical cofounders from the US and Mexico, who met in Switzerland. We believe very strongly in catering for a global learner base and integrating different learning cultures directly into the design.

In our process of building a company, we believe in hiring people with unique perspectives and always talking directly to the groups of users we are hoping to help. We believe a work environment, especially in a startup, should be a place where you feel welcome, comfortable to talk to anyone in the company, and able to bring all aspects of yourself to work. 

Sitting side by side in the research lab for the last four years has taught us a lot about how to build a culture where our colleagues are excited to come to work every day.

Scholé's business model focuses on providing tailored, efficient, and flexible upskilling in data science to adult learners, particularly those transitioning within or into tech-centric roles. The platform delivers two primary services: customized learning tracks (Goal-Oriented and Profession-Oriented) and a personalized curriculum development using AI-driven tools. This model addresses specific professional needs, making learning relevant and directly applicable, thus enhancing job performance and career development.

Key Customers and Beneficiaries

1. Adult Learners: Individuals seeking to enhance their data science skills for professional advancement or transition into new roles. They benefit from Scholé's personalized learning paths, which are designed to fit their specific career contexts and goals.

2. Employers and Corporations: Companies looking to upskill their workforce to stay competitive in a rapidly evolving market. Scholé aligns its curriculum with industry demands to ensure that learning outcomes meet current and future job market needs.

Products and Services

• Personalized Learning Tracks: Offering a choice between a goal-focused track for immediate problem-solving skills and a profession-oriented track for comprehensive career development in data science.

• Multilingual Learning Modules: Facilitating inclusive education by providing courses in multiple languages, thus reaching a global audience and breaking down language barriers in learning.

• AI-Tutoring and Curriculum Customization: Utilizing fine-tuned LLMs to develop custom problem sets and learning modules based on individual input and career requirements.

Value Proposition

Scholé provides value through its highly personalized and flexible learning solutions that are designed to be directly applicable to various professional scenarios. This makes the learning process more relevant and engaging but also ensures that the skills acquired are immediately useful in the learners’ professional lives. The platform’s ability to offer content in multiple languages and its focus on accessibility help in reaching a diverse global audience, making advanced education more inclusive.

Revenue Streams

1. Subscription Fees: Learners pay a subscription fee for access to personalized tracks and ongoing curriculum updates. Pricing models vary to accommodate different learning needs and economic backgrounds, including discounted rates for learners from underserved communities.

2. Partnership and Licensing Fees: Revenue is also generated through partnerships with educational institutions and companies that incorporate Scholé's platform into their training programs.

3. Grant Funding and Donations: Funding from educational grants and donations supports the platform’s accessibility initiatives and helps subsidize costs for users in need.

By collaborating with community organizations and tailoring its offerings to meet diverse needs, Scholé enhances educational equity and supports community development. Our commitment to ethical AI use and inclusive design further solidifies its role as a socially responsible educational platform.

Scholé's plan for achieving financial sustainability is built upon a multi-faceted revenue model that includes subscription fees, corporate partnerships, educational grants, and strategic investments. Below, we detail each component of our financial strategy:

1. Subscription Fees: Our primary revenue source will come from subscription fees charged to individual learners and corporate clients. We employ a freemium model: providing a single personalized learning plan for free to build user interest and adoption and a premium subscription fee for individuals and companies. Individuals can access our learning tracks and personalized curriculum on a monthly or yearly basis, while companies can subscribe at a corporate rate that provides access for teams or departments. This approach is scalable and provides a steady income stream. 

Additionally, we plan to launch an expert version of our platform that offers advanced features (e.g. knowledge tracing, recommending course content for employees) and analytics (understanding which skills your employees are upskilling on), targeting high-end corporate clients and institutions. This tiered pricing strategy (free, premium for individuals, premium for organizations) will allow us to cater to a broader range of customers while boosting our average revenue per user (ARPU).

2. Corporate Partnerships and Contracts: By established partnerships with various companies to integrate our platform into their ongoing professional development programs with multi-year and including both licensing fees and custom development charges.

3. Educational Grants and Funding: We are actively applying for and in the process of receiving educational grants that support the development and expansion of our AI learning programs. Joining the MIT Solve ecosystem would be a great initial step.