📶 Levels 📚 Courses 🗺 Learning Paths 🔬 Topics 🛠 Resources 📖 Glossary Join Now →
5 Levels · Basics to Live Projects & APIs

Learn Artificial Intelligence
the right way

From absolute zero to shipping real AI products — 5 structured levels, real projects, and a community of 50,000+ learners.

Start at Level 1 → Explore Paths
50K+
Active Learners
5
Skill Levels
120+
Courses
Free
To Join
// subject tracks

AI + Everything
You Already Love

Pick your domain. Learn how AI supercharges it. Build real projects at the intersection.

🌐
Track 01
Website Development + AI

Build AI-powered web apps — chatbots, smart search, personalized UX — with modern frameworks.

React Next.js LLM APIs
📚 8 courses · 24h
📱
Track 02
App Development + AI

Ship iOS & Android apps with on-device ML, voice assistants, and real-time AI features.

Flutter React Native CoreML
📚 7 courses · 20h
☁️
Track 03
Cloud Computing + AI

Deploy scalable AI services on AWS, GCP & Azure. Serverless inference, auto-scaling, MLOps.

AWS GCP Docker
📚 9 courses · 28h
🗄️
Track 04
Data Engineering + AI

Build AI-ready pipelines — ETL, feature stores, real-time streaming, and vector databases.

Spark Airflow dbt
📚 10 courses · 32h
🐍
Track 05
Python + AI

Go from Python basics to building production ML models, APIs, and automation tools with AI.

NumPy FastAPI PyTorch
📚 12 courses · 40h
📲
Track 06
Social Media Management + AI

Automate content, grow audiences, and analyze trends using AI writing and scheduling tools.

GPT APIs Analytics Automation
📚 6 courses · 18h
// curriculum

5 Levels to AI Mastery

A complete progression from zero knowledge to building real-world AI products with live APIs.

🌱

Level 1 — The Basics

No prior experience needed. Learn what AI actually is, set up your environment, and write your first ML code. You'll finish with a working classifier.

⏱ 4–6 weeks🐣 Beginner🐍 Python basics required
8
Modules
40h
Content
5
Projects
📘 What you'll learn
🧮
Math for AI
Vectors, matrices, dot products, basic calculus
📊
Statistics & Probability
Mean, variance, distributions, Bayes' theorem
🐍
Python for Data Science
NumPy, Pandas, Matplotlib — the essential toolkit
📈
Linear Regression
Predict values, understand loss & optimization
🔀
Logistic Regression
Binary classification, sigmoid, decision boundaries
🌲
Decision Trees
Gini impurity, info gain, overfitting prevention
🎯
Model Evaluation
Accuracy, precision, recall, F1, confusion matrix
🔧
Scikit-learn Basics
Pipelines, train/test split, cross-validation
🛠 Level 1 Projects
🏠 House Price Predictor

Linear regression on Boston housing data using Pandas and Scikit-learn.

Linear Regression
📧 Spam Email Classifier

Classify emails using logistic regression and bag-of-words features.

Classification
🌸 Iris Flower Classifier

Multi-class classification with decision trees. Visualize decision boundaries.

Decision Tree
📊 Data EDA Dashboard

Exploratory analysis on a real dataset — distributions, correlations, Seaborn.

EDA
🎯 Titanic Survival Prediction

Kaggle's starter — feature engineering, missing data, model comparison.

Kaggle
🔥

Level 2 — Basic to Moderate

Advance from simple models to ensemble methods, feature engineering, and your first neural networks. Bridge toy examples and real datasets.

⏱ 6–8 weeks📶 Basic→Moderate✅ Level 1 recommended
10
Modules
55h
Content
6
Projects
📘 What you'll learn
🌳
Random Forests
Bagging, feature importance, ensemble intuition
🚀
Gradient Boosting / XGBoost
Boosting, XGBoost, LightGBM, hyperparameter tuning
🔍
Feature Engineering
Encoding, scaling, imputation, polynomial features
📉
Dimensionality Reduction
PCA, t-SNE, UMAP — visualizing high-dim data
🔵
Clustering
K-Means, DBSCAN, hierarchical clustering
🧠
Intro to Neural Networks
Perceptrons, backpropagation, activation functions
⚙️
Keras & TensorFlow
Sequential API, Dense layers, training loops
🛡
Cross-Validation & Tuning
GridSearchCV, RandomSearch, overfitting control
📦
Sklearn Pipelines
Pipelines, ColumnTransformer, full workflow
📝
Text Classification Basics
TF-IDF, Naive Bayes, basic NLP preprocessing
🛠 Level 2 Projects
💳 Credit Card Fraud Detection

Imbalanced dataset — SMOTE, Random Forest and XGBoost with ROC/AUC.

Imbalanced Data
🎬 Movie Review Sentiment

IMDB reviews via TF-IDF + Logistic Regression, then a small neural net.

NLP Basics
🛒 Customer Segmentation

K-Means + PCA on retail data to segment customers by purchasing behavior.

Clustering
📈 Stock Price Classification

Predict up/down movement with XGBoost and engineered technical indicators.

Time Series
🏥 Diabetes Prediction

Medical classification with Keras MLP — handle missing values and class imbalance.

Healthcare AI
🏆 Kaggle Competition

Participate in a real intermediate Kaggle competition from EDA to submission.

Competition
⚡

Level 3 — Moderate to Advanced

Master deep learning: CNNs, RNNs, Transformers from scratch. Work with PyTorch, HuggingFace, and implement architectures used in production.

⏱ 8–10 weeks📶 Moderate→Advanced✅ Level 2 recommended
12
Modules
70h
Content
7
Projects
📘 What you'll learn
🔥
PyTorch Deep Dive
Tensors, autograd, custom datasets, DataLoaders
👁️
Convolutional Neural Nets
Conv2D, pooling, ResNet, VGG, EfficientNet
🔁
RNNs & LSTMs
Vanishing gradients, LSTM gates, GRU, sequences
🔮
Transformers from Scratch
Self-attention, multi-head, positional encoding
📝
BERT & GPT Architecture
Pre-training, tokenization, fine-tuning tasks
🎨
GANs
Generator/discriminator, DCGAN, training stability
📍
Object Detection
YOLO, SSD, anchor boxes, mAP evaluation
🗣
Speech & Audio AI
Spectrograms, Whisper, audio classification
📌
Transfer Learning
Feature extraction, fine-tuning pretrained models
⚡
Training at Scale
Mixed precision, gradient accumulation, multi-GPU
📦
HuggingFace Transformers
Pipelines, AutoTokenizer, Trainer API
🔬
Experiment Tracking
W&B, MLflow — log metrics, compare runs
🛠 Level 3 Projects
😷 COVID X-Ray Classifier

Fine-tune ResNet50 on chest X-rays — COVID/Normal/Pneumonia, 95%+ accuracy.

CNN + Transfer Learning
📰 News Category Classifier

Fine-tune BERT on BBC News for multi-class text classification via HuggingFace.

BERT Fine-tuning
🎵 Music Genre Classifier

Mel spectrogram + CNN to classify 10 genres from GTZAN in PyTorch.

Audio AI
🚗 Car Object Detection

Train YOLOv8 on custom vehicle dataset, deploy real-time webcam inference.

YOLO
🎭 Face GAN Generator

Train a DCGAN on CelebA to generate realistic faces from noise vectors.

GAN
📈 Time Series Forecasting

LSTM + Transformer hybrid for multi-step electricity demand forecasting.

LSTM + Transformer
🈲 Multilingual Sentiment

Fine-tune XLM-RoBERTa for sentiment across 5 languages — cross-lingual transfer.

Multilingual NLP
🚀

Level 4 — Advanced

Frontier AI engineering: LLMs, RAG pipelines, RL, diffusion models, MLOps, model optimization. Start building real production AI systems.

⏱ 10–12 weeks🔴 Advanced✅ Level 3 recommended
14
Modules
90h
Content
8
Projects
📘 What you'll learn
🤖
LLM Architecture Deep Dive
GPT internals, KV cache, attention, scaling laws
🎓
RLHF & Alignment
Reward modeling, PPO, DPO, Constitutional AI
🔎
RAG Systems
Chunking, embeddings, vector DBs, re-ranking
🕹
Reinforcement Learning
MDPs, Q-learning, DQN, PPO, Gym environments
🖼
Diffusion Models
DDPM, score matching, LoRA + Stable Diffusion
🤝
AI Agents & Tool Use
ReAct, function calling, LangChain, AutoGPT
💾
Vector Databases
FAISS, Pinecone, Weaviate, Qdrant, HNSW
⚗️
LoRA / QLoRA Fine-tuning
Parameter-efficient fine-tuning, adapters, quantization
☁️
MLOps & Deployment
Docker, FastAPI, Modal, HuggingFace Spaces
🏎
Model Optimization
Quantization, pruning, distillation, ONNX
🔐
AI Safety & Red-Teaming
Adversarial prompts, jailbreak defenses, bias audit
🌐
Multimodal Models
CLIP, LLaVA, vision-language pipelines
📊
LLM Evaluation
BLEU, ROUGE, MMLU, LM-Eval, human evals
🧪
Prompt Engineering Pro
CoT, ToT, DSPy, automatic prompt optimization
🛠 Level 4 Projects
📚 Private Doc Q&A (RAG)

RAG pipeline over 100+ PDFs with LangChain, Chroma, and Claude/OpenAI.

RAG + Vector DB
🕹 RL Game Agent

DQN agent that beats Atari Breakout from scratch in PyTorch + Gym.

Reinforcement Learning
🎨 Custom LoRA Fine-tuning

Fine-tune Stable Diffusion XL on your own style dataset using LoRA.

LoRA + Diffusion
🤖 AI Research Agent

Autonomous agent that searches arXiv, reads papers, writes summaries.

AI Agents
🩺 Medical QA with RAG

Fine-tune Mistral-7B with QLoRA on medical data + RAG over clinical guidelines.

QLoRA + RAG
📷 Vision + Language App

Multimodal app using LLaVA — upload images, ask questions, get AI descriptions.

Multimodal
⚡ Serve LLM at Scale

Quantize Llama-3 to 4-bit, serve with vLLM + FastAPI, deploy on Modal.com.

MLOps
🔍 Semantic Search Engine

Full semantic search over Wikipedia — FAISS, sentence-transformers, React UI.

Vector Search
🏗

Level 5 — Live Projects & API Integrations

Build and ship real products. Integrate with production AI APIs. Full-stack AI apps with auth, billing, monitoring, and real users.

⏱ Ongoing🏆 Professional✅ All levels recommended
10
APIs
12
Live Projects
∞
Impact
🔌 Production AI APIs You'll Integrate
Anthropic (Claude)

Messages API, tool use, vision, streaming. Build intelligent chat, code review, and reasoning apps.

client.messages.create( model="claude-sonnet-4-20250514", messages=[{"role":"user", "content": prompt}] )
OpenAI (GPT-4o / DALL-E 3)

Chat completions, function calling, embeddings, image generation and vision at production scale.

openai.chat.completions.create( model="gpt-4o", messages=messages, tools=tools )
Google Gemini

Gemini Pro/Flash — multimodal inputs, 1M token context, grounding with Google Search.

model = genai.GenerativeModel( "gemini-1.5-pro" ) response = model.generate_content( [image, prompt] )
HuggingFace Inference

Run 500K+ open-source models via API — no GPU. Deploy your own model endpoints.

from huggingface_hub import InferenceClient client = InferenceClient() result = client.text_generation( prompt, model="mistralai/Mistral-7B" )
Replicate

Run Stable Diffusion, Llama, Whisper and 10,000+ models via REST. Pay per second of compute.

import replicate output = replicate.run( "stability-ai/sdxl", input={"prompt": prompt} )
ElevenLabs (Voice AI)

Text-to-speech, voice cloning, real-time synthesis. Build AI podcasts and voice assistants.

audio = client.generate( text="Hello from AIstudyClub", voice="Rachel", model="eleven_turbo_v2" )
OpenAI Whisper API

99-language speech-to-text transcription and translation. Meeting transcribers and voice apps.

transcript = client.audio .transcriptions.create( model="whisper-1", file=audio_file )
Pinecone (Vector DB)

Managed vector database for semantic search and RAG at scale. Billions of vectors, ms queries.

index.upsert(vectors=[{ "id": "1", "values": embedding, "metadata": {"text": chunk} }]) results = index.query(top_k=5)
🚀 Capstone Live Projects
🤖 AI SaaS Chatbot Platform

Full-stack SaaS with Claude API backend, Next.js frontend, Stripe billing, Clerk auth, conversation history in Postgres. Deploy to Vercel with custom domain.

Claude APINext.jsStripePostgresVercel
📚 AI Study Assistant App

Upload PDFs, auto-generate flashcards & quizzes with GPT-4. RAG over study materials, spaced repetition tracking. Mobile-friendly React PWA with offline support.

OpenAI APIPineconeReactFastAPISupabase
🎨 AI Image Studio

Generate, edit, and upscale images using DALL-E 3, Stable Diffusion XL via Replicate, and Real-ESRGAN. User gallery, social sharing, credit system.

DALL-E 3ReplicateCloudinaryReact
🎙 AI Meeting Transcriber & Summarizer

Real-time transcription with Whisper API, speaker diarization, key point extraction with Claude, action items with Slack/Notion integration.

Whisper APIClaude APISlack APIWebSockets
🛒 AI E-commerce Recommendation Engine

Product semantic search with embeddings, personalized recommendations with collaborative filtering, LLM-generated descriptions at scale.

OpenAI EmbeddingsWeaviateFastAPIRedis
🎓 Adaptive AI Tutoring Platform

Gemini tutors students, adapts difficulty in real-time, generates practice problems, tracks mastery per topic. ElevenLabs voice for audio lessons.

Gemini APIElevenLabsNext.jsPrisma
🔍 AI Code Review Bot

GitHub App that reviews PRs with Claude, finds bugs, suggests improvements, generates docs. CI/CD integration, configurable rules per repository.

Claude APIGitHub AppsNode.jsDocker
📊 AI Data Analyst Dashboard

Natural language to SQL with GPT-4, auto-generate charts, anomaly detection, scheduled AI-written reports. Connect to PostgreSQL and BigQuery.

GPT-4oLangChainPlotlyBigQuery
// featured courses

Featured Courses

Hands-on, project-based courses designed by AI practitioners.

🧠
Beginner

Machine Learning Fundamentals

Core ML concepts — linear regression to decision trees — with hands-on Python labs.

⏱ 12h📚 32 lessons⭐ 4.9
🔥
Hot 🔥

Deep Learning with PyTorch

Build neural networks from scratch. CNNs, RNNs, Transformers — master modern AI architectures.

⏱ 20h📚 55 lessons⭐ 4.8
💬
Intermediate

Natural Language Processing

Tokenization, embeddings, BERT, GPT fine-tuning. Build chatbots and text classifiers.

⏱ 16h📚 44 lessons⭐ 4.7
👁️
Intermediate

Computer Vision Mastery

Object detection, image segmentation, GANs. Build apps that see with OpenCV and YOLO.

⏱ 18h📚 48 lessons⭐ 4.8
🤖
Advanced

Prompt Engineering & LLM APIs

Master Claude, GPT-4, Gemini APIs. Chain-of-thought, RAG, agents, and tool use.

⏱ 10h📚 28 lessons⭐ 4.9
🎨
New

Generative AI Applications

Build with Stable Diffusion, DALL-E, Midjourney APIs — image, video, and audio generation.

⏱ 14h📚 38 lessons⭐ 4.9
// learning paths

Choose Your Path

Structured roadmaps to take you from curious beginner to confident AI practitioner.

01

🌱 AI Beginner Path

Zero to first working ML model in 4 weeks. No math degree required.

  • Python for Data Science (5h)
  • Math Essentials for AI (4h)
  • ML Fundamentals (12h)
  • Scikit-Learn Projects (6h)
  • Capstone: Kaggle Competition
02

🚀 AI Engineer Path

Go from ML basics to deploying production AI systems in 3 months.

  • Deep Learning with PyTorch (20h)
  • MLOps & Model Deployment (10h)
  • LLM APIs & Prompt Engineering (10h)
  • Vector Databases & RAG (8h)
  • Capstone: AI SaaS Product
03

🔬 AI Researcher Path

For those who want to understand and advance the frontier of AI.

  • Advanced Deep Learning (25h)
  • Transformers from Scratch (15h)
  • Reading AI Papers (8h)
  • Reinforcement Learning (18h)
  • Capstone: Original Research
// all topics

Explore AI Topics

Browse every subject — from foundations to the bleeding edge.

🧮Linear Algebra
📊Statistics
🐍Python for ML
🌲Decision Trees
📉Gradient Descent
🧠Neural Networks
🔁RNNs & LSTMs
👁️CNNs
🔮Transformers
💬LLMs & GPT
🎨Diffusion Models
🕹️Reinforcement RL
🗣️Speech AI
🔎Semantic Search
📦Vector Databases
⚙️MLOps
☁️AI on Cloud
🏎️Model Optimization
🔐AI Safety
🤝AI Agents
// resources

Free Learning Resources

Curated books, papers, tools, and cheatsheets to accelerate your learning.

📄 Paper

Attention Is All You Need

The 2017 landmark paper that introduced the Transformer — backbone of all modern LLMs.

Read on arXiv →
📚 Book

Deep Learning (Goodfellow et al.)

The definitive DL textbook. Covers foundations through advanced topics. Freely available.

Read Free →
🛠 Tool

HuggingFace Hub

The GitHub of AI — 500,000+ models, datasets, and apps. Essential for every practitioner.

Explore Hub →
🎓 Course

Fast.ai Practical Deep Learning

Top-down practical approach by Jeremy Howard. Start coding on day one.

Take Course →
📝 Cheatsheet

ML Algorithm Cheatsheet

Quick reference for choosing the right algorithm, key hyperparameters, and when to use what.

Download PDF →
🎯 Playground

TensorFlow Playground

Interactive neural network visualizer. Experiment with architectures in your browser.

Open Playground →
// glossary

AI Terms Explained

Plain-English definitions for concepts you'll encounter on your AI journey.

Transformer
An attention-based neural network that processes sequences in parallel — backbone of GPT, BERT, and all modern language models.
Gradient Descent
The optimization algorithm adjusting model weights by moving in the direction of steepest decrease in loss, enabling models to learn from data.
Embedding
A dense vector representation where similar items are geometrically close — used for words, images, users, and more.
Fine-tuning
Training a pre-trained model on a smaller task-specific dataset so it adapts its general knowledge to a particular downstream task.
RAG
Retrieval-Augmented Generation — combines a retrieval system with an LLM to ground answers in external documents or databases.
Attention Mechanism
A neural component that lets a model weigh the relevance of different input parts when producing each output token.
Overfitting
When a model memorizes training data too closely, performing well on training examples but poorly on new, unseen data.
Diffusion Model
A generative model that learns to reverse a noise-adding process, enabling high-quality image generation (Stable Diffusion).
Reinforcement Learning
A learning paradigm where an agent learns by taking actions, receiving rewards or penalties, and optimizing for max cumulative reward.
Prompt Engineering
Crafting effective inputs to guide language models — using few-shot examples, chain-of-thought, and structured instructions.
Vector Database
A database optimized for storing and retrieving high-dimensional embeddings via similarity search — essential for RAG systems.
LoRA
Low-Rank Adaptation — trains small adapter matrices instead of the full model, cutting GPU cost by 10–100× for fine-tuning LLMs.

Stay ahead of the AI curve

Weekly digest: new courses, cutting-edge papers, job opportunities, and community highlights. No spam, ever.