Personalized Learning Paths

The Differentiation Dilemma

Sarah teaches 7th-grade Math in a public middle school in Colorado. She has 28 students. Six of them are still shaky on 5th-grade concepts — they struggle with basic fractions. Sixteen are roughly at grade level. Four are bored because they finished the chapter two days ago. Two have IEPs that require modified assignments and extended time. Sarah knows she should differentiate instruction. She also knows she has 50 minutes per period, a shared Chromebook cart, and no teaching assistant.

This is not a hypothetical. This is the reality for millions of teachers across the US, UK, and Australia. AI cannot solve the class-size problem, but it can do something powerful: it can help you create three learning paths in the time it used to take to create one.

What Personalized Learning Actually Means

Let us clear up a common misconception. Personalized learning does not mean every student gets a unique curriculum. That is neither practical nor necessary. It means you systematically address three tiers — which aligns directly with MTSS (Multi-Tiered System of Supports) and RTI (Response to Intervention) frameworks already used in most US districts:

In the UK, this maps roughly to differentiation within the National Curriculum — support, core, and stretch activities within each Key Stage. In Australia, it aligns with the Nationally Consistent Collection of Data (NCCD) framework for students with disability and the general capability continuum in the Australian Curriculum.

Identifying Learning Gaps from Data

Before you can personalize, you need to know where each student stands. Most districts already have data from diagnostic assessments (MAP, iReady, Star), state testing, or classroom formatives. If you do not have formal data, a simple spreadsheet works.

Look at data/student-performance.csv — it contains anonymized data for a fictional Grade 7 section of 28 students across 5 math topics. Each cell shows whether the student scored Below Expectations (B), Meeting Expectations (M), or Exceeding Expectations (E) on each topic.

Reading the Patterns

When you look at this data, several patterns emerge:

You can ask AI to analyze patterns like these:

Here is my class performance data for 5 topics (B=Below, M=Meeting, E=Exceeding):

[paste the data]

Identify:
1. Students who need Tier 3 intervention (3+ topics at B level)
2. Topics where more than 40% of students scored B
3. Prerequisite gaps — which earlier topics predict failure on later ones
4. Students ready for extension activities (3+ topics at E level)

AI cannot replace your judgment about why a student is struggling — maybe Emma missed two weeks because of a family move, or Jayden's parents are going through a divorce. But it can flag the patterns faster than you can spot them manually in a spreadsheet.

Creating Personalized Paths

Once you know who needs what, here is how to generate differentiated materials efficiently.

For Tier 3 Students: Scaffold and Simplify

Marcus is in Grade 7 but his fraction skills are at a Grade 5 level. He has an IEP that specifies simplified language and visual supports.

Prompt:

"Create a 5-day remedial plan for a Grade 7 student whose fraction skills are at Grade 5 level. The goal is to bring them up to Grade 6 level so they can participate in the current unit on percentages. Each day: one concept, one worked example with visual models (area models and number lines), 5 practice problems with answers. Use a Lexile level of 700-800. Include notes on where manipulatives (fraction tiles, fraction strips) would be helpful."

For Core Students: Reinforce and Practice

The middle group needs solid practice with variety. Same concepts, different contexts.

Prompt:

"Generate 15 practice problems on converting fractions to percentages (Grade 7, CCSS 7.RP.A.3). Mix of: 5 straightforward calculations, 5 word problems using real-world contexts (basketball free-throw percentages, store discounts, recipe scaling), 5 that require two steps. Include answer key."

For Extension Students: Extend and Challenge

Olivia finishes every worksheet in 10 minutes and then distracts her neighbors. She needs challenge, not more of the same.

Prompt:

"Create 5 challenge problems on percentages for an advanced Grade 7 student. Include: one involving successive discounts (a store offers 20% off, then an additional 10% — is that the same as 30% off?), one involving percentage increase and decrease, one real-world data interpretation problem using census or economic data. These should require multi-step reasoning. Do not include answers — the student should present their solutions to the class."

The Learning Path Document

For each unit, create a one-page learning path that you share with students (or upload to your LMS). Here is a template:

Check data/learning-paths.json for six pre-built path templates across Math, Science, and ELA for Grades 6-10. Each path includes the prompts you need to generate the materials.

IEP and 504 Accommodations

AI is particularly valuable for creating accommodated materials efficiently. Common accommodations and how AI helps:

Always verify that AI-generated accommodated materials actually match the IEP or 504 plan. A simplified version that removes the assessed skill entirely is not an accommodation — it is a different assignment.

Practical Constraints and Workarounds

"I Have 28 Students and 50 Minutes"

You cannot run three parallel classes. But you can:

"My District Has Strict Technology Policies"

If your district restricts which AI tools teachers can use, focus on using AI for preparation, not live instruction. Generate materials at home, review them, and bring the finished product to class. The students never interact with AI directly — you do.

"Some Students Have Significant Learning Disabilities"

For students with more intensive needs, AI can help you create multi-modal materials:

Prompt:

"Convert this 10-question written math quiz into three formats: (1) a standard written version, (2) a version with visual models and sentence starters for each question, and (3) instructions for an oral assessment where the teacher reads problems aloud and the student uses manipulatives to show their thinking."

Tracking Progress Over Time

The real power of personalized paths is not a single differentiated worksheet — it is tracking whether the intervention is working. Keep a simple tracker (many districts have this in their data management system, but a spreadsheet works too):

When Jayden shows no improvement after six weeks of Tier 1 and Tier 2 intervention, that is data you bring to the SST or IEP team. You are not guessing — you have documented the gap, the intervention, and the result. This is exactly what MTSS frameworks are designed for.

Key Takeaways