TechPath Learning Intelligence

What can TechPath do?

TechPath builds a personalized, dependency-aware learning roadmap from across the open web in a single tool: real courses, tutorials, and videos for your goal, staged from your current level to interview-ready, plus quizzes, projects, and matched opportunities. Add a goal with your current level and you can:

📦 What you can build	⚙️ Features & integrations
🗺️ Staged roadmaps from beginner foundations to advanced, interview-ready skill	🔎 Live discovery across search results and YouTube, not a static curated list
📝 Quizzes per stage to check you're actually ready to move on	📅 Freshness-aware — every resource is discovered at run time, not hardcoded
🛠️ Projects and a capstone brief — problem statement, milestones, rubric	⬇️ Export data to JSON, CSV, or Excel
🎯 Opportunity matches — hackathons, internships fit to your current stage, not just the most popular one	🔌 API access via the Apify API and CLI
📄 Human-readable report as well as structured data	🤖 LLM-structured output, built for MCP, Claude, Gemini, or ChatGPT to consume downstream

If you're a beginner without a mentor, an intermediate learner who doesn't know what's next, a career switcher, a student prepping for interviews, or an educator building a curriculum, you can get a full personalized path from a single tool.

What data can I get from TechPath?

TechPath is the structure the open web is missing — the roadmap, the proof-of-readiness, and the next real opportunity behind any tech-learning goal:

• Roadmap stages: what to learn, and in what order. Foundations → Core Skills → Specialization → Build → Prove → Advance, each with the specific skills that stage covers.
• Resources: real, live links — not a stale list. Articles, docs, and videos discovered from the web at the moment you run it, each with its title, URL, and type.
• Quizzes: proof you're ready to move on. A short skill check per stage — conceptual questions, coding prompts, or mini tasks.
• Projects: proof you can build it. A practical project per stage, plus a full capstone brief — problem statement, objectives, stack, milestones, evaluation rubric — for your target role.
• Opportunity matches: what's actually next for you. Hackathons, internships, and competitions matched to your current stage, with the reason each one fits.

You can generate a roadmap from just a goal and level, or refine it with your known skills, hours per week, and interests for a more personalized path.

How It Works — Step by Step

1. You submit a learner profile

You fill in your goal, currentLevel, and — optionally — your knownSkills, hoursPerWeek, and interests. Apify auto-generates this form directly from the input schema, so no coding is needed to try it from the Console.

2. TechPath calls other Actors to discover real resources

Once you click Start, TechPath's own code (main.py) runs on Apify's servers. The first thing it does is call two other Actors already published on the Apify Store, as sub-runs:

• apify/google-search-scraper — searches the live web for courses, tutorials, and roadmap content matching your goal
• streamers/youtube-scraper — searches YouTube for relevant tutorial videos

This is genuine Actor-to-Actor composition, not a single monolithic scraper: TechPath's code calls Actor.call(actor_id=..., run_input=...), which starts each of those Actors as its own independent run, waits for it to finish, and pulls its dataset back in. Nothing here is hardcoded — every resource returned is whatever those Actors find live, at the exact moment you run TechPath.

Screenshot to insert: either the main.py source showing the Actor.call() lines, or the Console "Runs" view showing the nested google-search-scraper and youtube-scraper runs kicking off underneath your TechPath run.

3. The LLM structures the raw data into a roadmap

Raw scraped titles, URLs, and descriptions are just a messy list on their own — not a roadmap. TechPath sends that raw data, together with your profile, to an LLM with strict instructions: build a staged roadmap (Foundations → Core Skills → Specialization → Build → Prove → Advance), generate a short quiz and a practical project for every stage, and only ever reference resources that actually appeared in the scraped data — never invent a URL. The LLM returns this as one structured JSON object, which TechPath validates before using it.

4. Results are saved two ways

Each roadmap stage is pushed to the Actor's dataset as its own row — one Actor.push_data() call per stage — giving you structured, machine-readable output. At the same time, TechPath renders a styled HTML version of the same roadmap and saves it to the key-value store under the key REPORT, so a non-technical learner can open one link and read a clean page instead of raw JSON.

*Screenshot to insert: the Storage → Dataset tab, showing the actual roadmap-stage rows TechPath returned for a real

5. You're charged per what was actually produced

Because every dataset row comes from a real push_data() call, turning on the apify-default-dataset-item synthetic PPE event means every roadmap stage automatically generates a small charge — no separate billing code anywhere in main.py. Turning on apify-actor-start charges once per run, the moment it begins.

Mode 1: Quick roadmap

Enter your target role and current level, and TechPath builds a full staged roadmap using sensible defaults for hours per week and interests.

• goal, e.g. "Machine Learning Engineer"
• currentLevel: beginner, intermediate, or advanced

Mode 2: Personalized deep roadmap

Add your known skills, hours per week, and interests, and TechPath tailors resource selection, pacing, and the capstone project to you specifically — e.g. an ML roadmap that leans into healthcare AI because that's what you told it you care about.

Add this field	And you get
knownSkills	Skill gaps calculated against your actual starting point, not a generic beginner
hoursPerWeek	A pace-aware roadmap rather than a one-size-fits-all timeline
interests	Specialization stages and capstone project themed to what you actually want to build

Output and setup

Results land in a dataset — one row per roadmap stage. View it as a table in the Storage tab, download as JSON, CSV, or Excel, or pull it via the API. A parallel human-readable HTML report is saved to the key-value store under REPORT, so non-technical learners never have to read raw JSON.

Dataset sample:

{
  "stage": "Core ML",
  "skills": ["supervised learning", "scikit-learn"],
  "resources": [
    {"title": "Intro to Scikit-learn", "url": "https://...", "type": "article"}
  ],
  "quiz": [
    {"question": "What is overfitting?", "answer": "..."}
  ],
  "project": "Build and evaluate a classification model on a public dataset"
}

How much will TechPath cost?

TechPath uses a pay-per-event (PPE) pricing model — you're charged for what the run actually produces, not a flat run fee:

Event	What it means
apify-actor-start	Charged once when your roadmap run starts
apify-default-dataset-item	Charged per roadmap stage returned

Team

Udotong Peace

Mayowa Titilola

Obiajulum Ndubuisi

Track: EduTech — Access to education and scholarships Built for: Apify x She Code Africa BuildHer Hackathon 2026 — Ship and Earn Africa