Lucid

AI that teaches you, guides you, and finds you work.

Comment

Inspiration

Zimbabwean youth want digital skills, but most learning platforms assume prior knowledge. When a student tries to learn coding, data science, or AI, they often get lost because the explanations are not adapted to their level. Many drop out before they gain confidence.

At the same time, Zimbabwe has a fast growing demand for digital workers in AI, fintech, e commerce, data analytics, and green tech. There is a clear gap between the desire to learn and the ability to access structured guidance.

Lucid was inspired by this problem. I wanted to build a system that adapts to the learner, understands how they learn, teaches them clearly, and then connects them directly to real earning opportunities. The goal was to create a complete pipeline, from skills to opportunities to income, supported by the Econet ecosystem.

What it does

  • Provides personalized explanations based on the user's learning style.
  • Converts textbooks and PDFs into clear lessons with visuals and voice.
  • Connects to Akello Smart Learning and adapts the content to the individual.
  • Allows easy subscription through USSD *143# on an Econet line.
  • Gives users a voucher or code to unlock Lucid premium.
  • After teaching, it recommends job or gig opportunities that match the user's skills.
  • Payments go directly to the user's EcoCash wallet.

Lucid creates a full cycle: learn → master → work → earn.

How we built it

  • Designed a multi agent AI workflow where each agent focuses on a specific task, for example understanding documents, generating explanations, analyzing the learner profile, and matching jobs.
  • Used Node.js for backend logic to handle user sessions, learning history, voucher redemption, and EcoCash flow.
  • Connected Lucid to Akello Smart Learning so the AI can read books and course materials and then convert them into personalized lessons.
  • Built a clean interface that supports voice teaching, visual boards, and guided explanations.
  • Created a USSD friendly subscription process that allows a user to dial *143#, buy Lucid bundles, receive a voucher, and activate premium.

Challenges we ran into

  • Different learners have different backgrounds, so building explanations that work for both beginners and intermediate users was difficult.
  • Limited connectivity in some areas required us to design for low bandwidth.
  • Integrating with telecom services and payment systems required a lot of testing because of edge cases.
  • Ensuring that adaptive learning stays consistent without becoming confusing.
  • Connecting users to verified job opportunities required building new processes and partnerships.

Accomplishments that we are proud of

  • Built a working prototype that demonstrates the end to end user journey.
  • Implemented personalized learning sessions that adapt to how the user understands concepts.
  • Achieved integration with Akello Smart Learning for real content ingestion.
  • Designed a subscription flow that works with Econet and EcoCash.
  • Established Lucid as more than a learning tool. It is a pathway from skills to income.

What we learned

  • Real impact in Zimbabwe requires solutions that consider internet limitations, devices, and payment systems.
  • Adaptive learning can significantly reduce user frustration when the AI starts from the user's current understanding.
  • Integration across multiple systems is possible with the right architecture.
  • Moving from prototype to deployment requires user trust, accessibility, and stable performance.
  • Education and employment need to be combined. Skills alone are not enough. Youth want real opportunities.

What is next for Lucid

  • Add more skill tracks, for example coding, AI fundamentals, digital marketing, data analytics, and green tech skills.
  • Add offline or low bandwidth learning modes.
  • Build a verified job marketplace with ratings, employer validation, and EcoCash payouts.
  • Run pilot programs in universities and rural communities.
  • Form partnerships with Econet, schools, NGOs, and youth hubs.
  • Add voice based conversational learning and real time question answering.
  • Refine our multi agent architecture using more advanced models and better user profiling.

Built With

Share this project:

Updates