# How the Find-Your-Level Placement Quiz Maps Knowledge to a Starting Phase

> Discover how the Find-Your-Level placement quiz maps AI ML knowledge to your perfect starting phase. Get a personalized learning path with Skip, Review, or Do statuses.

- Repository: [Rohit Ghumare/ai-engineering-from-scratch](https://github.com/rohitg00/ai-engineering-from-scratch)
- Tags: how-to-guide
- Published: 2026-08-29

---

**The find-your-level placement quiz evaluates AI/ML competency across five domains with ten questions to produce a 0-10 score that maps directly to entry phases 1, 3, 7, 11, or 14 of the 511-lesson curriculum, then generates a personalized learning path with Skip, Review, or Do statuses for each phase.**

The `find-your-level` skill in the **rohitg00/ai-engineering-from-scratch** repository eliminates curriculum onboarding guesswork by dynamically placing learners at the correct starting phase based on existing knowledge. Unlike static course catalogs, this placement quiz parses canonical phase data from [`ROADMAP.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/ROADMAP.md) and applies deterministic scoring logic defined in [`skills/find-your-level/SKILL.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/skills/find-your-level/SKILL.md) to create an adaptive entry point.

## Ten-Question Assessment Structure

The quiz architecture treats placement as a binary classification problem across five verticals, aggregating subdomain scores into a total that selects the optimal curriculum entry point.

### Five-Domain Knowledge Coverage

The assessment spans **5 core knowledge areas** with **2 questions per domain**, yielding a maximum score of **10 points** (1 point per correct answer). The domains evaluated are:

- **Math & Statistics**
- **Classical ML**
- **Deep Learning**
- **NLP & Transformers**
- **Applied AI**

This structure ensures that a learner’s strength in one area (e.g., mathematics) does not mask weaknesses in another (e.g., transformer architectures), providing a granular profile of preparedness.

### Round-Based Scoring System

Rather than revealing a cumulative total only at the end, the quiz operates in **five rounds** (one per domain). After each pair of questions, the learner receives an **area score** (0-2). The overall score is computed only after the fifth round, as outlined in [`skills/find-your-level/SKILL.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/skills/find-your-level/SKILL.md) (lines 27-30).

## Score-to-Phase Mapping Logic

The core algorithm translates the 0-10 aggregate score into a concrete phase entry point within the 20-phase roadmap. This mapping is hardcoded in the skill definition to ensure consistency across all learner assessments.

### Entry Point Determination

According to [`skills/find-your-level/SKILL.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/skills/find-your-level/SKILL.md) (lines 73-81), the scoring tiers map as follows:

| Total Score | Entry Point Phase | Curriculum Meaning |
|-------------|-------------------|-------------------|
| 0-3 | **Phase 1** – Math Foundations | Begin with ground-up basics (~23 hours) |
| 4-5 | **Phase 3** – Deep Learning Core | Skip math fundamentals; start with DL prerequisites |
| 6-7 | **Phase 7** – Transformers Deep Dive | Bypass classical ML; focus on attention mechanisms |
| 8-9 | **Phase 11** – LLM Engineering | Strong foundations; begin building LLM applications |
| 10 | **Phase 14** – Agent Engineering | Mastered curriculum; jump directly to agentic systems |

This deterministic table ensures that a learner scoring 6/10 enters at **Phase 7**, bypassing 132 lessons of preliminary material while retaining access to prerequisite content for review.

## Personalized Learning Path Generation

After determining the entry phase, the skill constructs a comprehensive markdown table covering all 20 phases, assigning granular study recommendations based on per-domain performance.

### Skip, Review, and Do Statuses

The status logic (defined in [`skills/find-your-level/SKILL.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/skills/find-your-level/SKILL.md), lines 85-99) categorizes each phase into one of three states:

- **Skip** – Phases fully covered by the learner’s demonstrated competency. Hours display as `--`.
- **Review** – Phases where the learner scored **1/2** in the corresponding domain (partial knowledge). The phase is flagged for a quick refresher.
- **Do** – Phases requiring in-depth study, assigned to domains scoring **0/2** or to all phases at and above the entry point.

For example, a 1/2 score in *Math & Statistics* yields a **Review** flag for Phase 1 (line 92), acknowledging existing algebra knowledge while recommending verification of calculus prerequisites.

### Dynamic Hour Estimation

The skill **does not hard-code hour estimates**. Instead, it parses parenthesized hour counts directly from [`ROADMAP.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/ROADMAP.md) phase headings (e.g., `Phase 1: Math Foundations — ✅ (~23 hours)` on line 28). This synchronization ensures that updates to the curriculum’s time estimates in [`ROADMAP.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/ROADMAP.md) automatically propagate to generated learning paths (see lines 98-101 of [`SKILL.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/SKILL.md)).

## Integration with the AI Engineering Curriculum

The placement quiz functions as a front-end to the canonical curriculum defined in [`ROADMAP.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/ROADMAP.md). While [`README.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/README.md) lists the quiz as a ten-question assessment and [`site/lesson.html`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/site/lesson.html) exposes the `/find-your-level` command for agents, the skill’s authority derives from its direct parsing of the roadmap’s phase titles and hour metadata.

## End-to-End Quiz Execution Flow

The complete pipeline from invocation to personalized plan follows these six steps:

1. **Agent Presentation** – The agent presents ten multiple-choice questions via the `/find-your-level` command interface.
2. **Answer Scoring** – Learner responses are evaluated, producing a 0-10 aggregate score and five domain sub-scores.
3. **Entry Point Selection** – The total score selects the phase via the *Score-to-Entry-Point Mapping* table.
4. **Roadmap Synchronization** – The skill fetches current hour estimates from [`ROADMAP.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/ROADMAP.md) for all 20 phases.
5. **Table Generation** – A markdown table renders with **Skip**, **Review**, and **Do** statuses and cumulative hour totals.
6. **Learner Recommendation** – The learner receives a concise summary ("Your personalized path: ~X hours across Y phases") with prompts to persist the plan (`/start-learning`) or begin immediately (`/learn`).

```python

# Pseudo-code demonstrating agent invocation

result = agent.run_skill("find-your-level")
score = result.total_score          # Integer 0-10

entry_phase = result.entry_point   # "Phase 7"

learning_path = result.markdown_table  # Markdown table string

print(learning_path)

```

## Summary

- The **find-your-level** placement quiz uses a **10-question, 5-domain assessment** to calculate a granular 0-10 competency score.
- Scores map deterministically to **entry phases 1, 3, 7, 11, or 14** based on logic defined in [`skills/find-your-level/SKILL.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/skills/find-your-level/SKILL.md) (lines 73-81).
- The system generates a **personalized 20-phase roadmap** with **Skip**, **Review**, or **Do** statuses based on per-domain performance (lines 85-99).
- Hour estimates are **dynamically fetched** from [`ROADMAP.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/ROADMAP.md) to ensure curriculum synchronization (lines 98-101).
- Learners bypass irrelevant material while retaining the option to review partially known domains, optimizing the path through 511 lessons.

## Frequently Asked Questions

### What knowledge domains does the placement quiz cover?

The quiz evaluates five domains: **Math & Statistics**, **Classical ML**, **Deep Learning**, **NLP & Transformers**, and **Applied AI**. Each domain contributes 2 questions to the 10-question assessment, allowing the system to detect specific knowledge gaps rather than relying on a single aggregate metric.

### How does the quiz determine whether to skip or review a phase?

The system assigns **Skip** to phases where the learner scored 2/2 in the corresponding domain, **Review** where they scored 1/2 (indicating partial knowledge), and **Do** for 0/2 scores or any phase at/above the calculated entry point. This logic is implemented in [`skills/find-your-level/SKILL.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/skills/find-your-level/SKILL.md) between lines 85-99.

### Where does the find-your-level skill get its hour estimates?

Rather than using static values, the skill parses hour counts directly from [`ROADMAP.md`](https://github.com/rohitg00/ai-engineering-from-scratch/blob/main/ROADMAP.md) phase headings (e.g., `(~23 hours)`). This ensures that any curriculum updates to time estimates automatically reflect in generated learning paths without modifying the skill definition.

### What happens if I score a perfect 10 on the placement quiz?

A perfect score of **10** maps to **Phase 14 – Agent Engineering**, allowing learners to bypass the first 13 phases (covering math, classical ML, deep learning, and transformers) and begin immediately with agentic AI systems. The learning path will mark all prior phases as **Skip**, though learners can still choose to review specific topics.