Most of the language learning apps focus on teaching an entire language - however the users on such apps are a mix of those who actually want to learn a language as a primary goal and those who want to learn it so they could travel, work, or have a specific type of conversation with locals.
Language learning is complex, we can avoid the complexity for users for whom learning that language is itself a secondary goal. Hence the idea of micro learning for languages. Want to travel to Japan? - learn enough of the language that would help you with a great travel experience. Want to negotiate groceries in Kannada? - then the user does not need to learn about family, ordering food, etc.
Language learning is just one such use case. The need for micro learning is omnipresent for all non specialists in a particular field - you want to implement some devops as a frontend engineer, you want to understand a prescription as a patient, you want to understand some economics as a politician, and so on.
Achieving top 5% speciality in a field takes years of efforts for which degrees are suitable. Investing time in a speciality curriculum is impractical when your goal is very specific requiring generalist know-how. Hence for such users, we need tools that are built to get the user’s job done! No traversing the depth and breadth of the ocean for every single task, instead learn accurately what you wish to practice and achieve your goal.
User Flow
A happy flow between the user and the product (mentioned as agent) is visualised in the following conversation,
User: “I need to negotiate my yearly rental agreement with the Kannada speaking owner. My goal is to limit the rental increase below 10% while keeping the deposit same.”
Agent: “Let’s narrow down the goal even further. Please answer the following questions so that I could plan a micro-curriculum for you.
Q1. Describe the owner - are they male/female and how much older than you. Single person or multiple people are discussing the rental agreement? Q2. What does the current rental agreement state as facts - provide the total monthly rent, deposit, maintenance, and other relevant information. Q3. What are your arguments for limiting the rental increase? What are the owner’s arguments that you are aware of?”
User answers the questions
Agent: “Based on our discussion, we have the following summary and goal of the task. Let me design a micro curriculum that will help you communicate better with the owner.”
Agent: “Chapter 1: Basic Greetings in Kannada; Chapter 2: Vocabulary related to your situation; …”
Agent: “Here is your daily learning plan to understand words, phrases, and basic sentences that would help you negotiate your rent. We will notify you daily with a 15 min exercise that would help you achieve your micro learning goal.
If you complete the entire curriculum as per plan and a basic assessment, you would receive a skill certificate.”
User: “I commit to these learning goals! However I have a slight modification on the rental increment I wish to pitch.”
Agent: “Your curriculum has been modified as per updated goal. Enjoy learning!”
Architectural Requirements
We need an agentic flow with following roles,
- Input: Understand user input requirements and make them more accurate through conversation with the user. Converge onto a specific goal.
- Material: Collate and create learning content and condense it into a micro curriculum.
- Structure: Plan daily learning goals based on the curriculum and enable user to practice them
- Feedback: Assess user inputs and share feedback
User Experience
I believe its best to test (mvp) this on a conversational tool that the vast market is already comfortable with - Whatsapp as UX with templated message flow helping the user practice daily.
The best way to proceed with an mvp would be to use a tool that directly helps with whatsapp configuration instead of doing all the elements separately - this will help us focus on perfecting and testing the core - the micro learning agentic flow.