What is the Learning Agent?
The Learning Agent is a dedicated AI inside SolversBoard that operates at a higher level than Meg. While Meg focuses on the question in front of you right now, the Learning Agent zooms out and looks at the bigger picture: your performance over time, your recurring mistakes, your strengths, and where your time is best spent.
It is accessible from the sidebar and can be opened at any point during your session - or after, when you want to reflect on your progress.
What the Learning Agent does
1. Performance pattern analysis
The Agent reviews your session history and identifies patterns that aren't obvious question by question. It might notice:
- You consistently make sign errors in algebra but not in arithmetic
- You get geometry questions right on the first attempt but struggle with trigonometry
- Your accuracy drops significantly in the last 15 minutes of a session (suggesting fatigue)
- You answer easy questions correctly but abandon medium-difficulty questions prematurely
These are the kinds of insights that a human tutor builds over weeks of one-on-one sessions. The Learning Agent surfaces them after a few hours of practice.
2. Knowledge gap mapping
Based on your answer history, the Agent builds a map of your knowledge across the topics in your chosen subject. Topics are marked as:
- Strong - consistent correct answers, including on hard questions
- Developing - mixed results, suggesting partial understanding
- Gap - repeated errors or avoidance, suggesting the concept needs direct attention
This map updates after every session and is the basis for the Agent's study recommendations.
3. Personalised study plans
Ask the Agent "what should I study this week?" and it generates a structured plan based on your gaps, your upcoming exam (if you've set one), and how much time you have available.
A typical plan might look like:
> "You have 8 days before your WAEC Chemistry paper. Your strong areas - stoichiometry and atomic structure - need maintenance only (30 minutes each this week). Your gap areas - equilibrium constants and electrochemistry - need focused drilling: 45 minutes each on days 1, 3, and 5. I'll generate questions from those topics specifically."
The plan is not generic. It is built from your data.
4. Session debriefs
At the end of a session - or whenever you ask - the Agent can give you a debrief:
- How many questions you attempted and your accuracy
- Which sub-topics came up and how you performed on each
- What the session revealed about your understanding
- What to focus on next time
This is the "game film review" moment. Most students finish a study session and immediately forget what they struggled with. The Agent makes sure that information is captured and acted on.
5. Motivational coaching
The Learning Agent is aware of your streaks, your recent trajectory (improving or plateauing), and the distance between where you are now and your goal. It uses this to give you honest, constructive motivation - not empty praise.
If you've been on a three-day streak, it acknowledges it and explains what that consistency is doing to your retention. If you've been avoiding a hard topic, it names it directly and helps you break it into manageable pieces.
How to use it
Open the Learning Agent from the robot icon in the sidebar. You can ask it anything:
- "What are my weakest topics in maths right now?"
- "Build me a 2-week revision plan for JAMB Biology"
- "Why do I keep getting equilibrium questions wrong?"
- "How has my accuracy changed over the last 7 days?"
- "What should I practise next?"
It responds in plain language, with specific references to your actual performance data. It can also generate a question set targeted at any topic it identifies as a gap - so you can move directly from insight to practice.
The difference between Meg and the Learning Agent
Meg is in the room with you, question by question. She reads your working, gives immediate feedback, and coaches you through each problem.
The Learning Agent is your strategist. It operates at the session and week level. It sees the forest, not just the trees.
Used together, they give you both the micro-level coaching (Meg, per question) and the macro-level strategy (Learning Agent, per week) that characterise the best human tutoring relationships.
Available on all plans
The Learning Agent is included on all SolversBoard plans - Individual, Family, and School. On Family and School plans, parents and teachers can also request a Learning Agent summary for any learner in their group, giving them a clear view of each student's progress and needs without having to sit through every session themselves.
The Learning Agent is built on the same principle as everything we make: intelligence should make you more capable, not more dependent. Every insight it gives you is designed to help you study smarter - not to study for you.