Learning Path¶

This file is the curriculum. Finish Part 1 first, then pick one or two tracks from Part 2 based on the roles you want.

Before You Start¶

Read genai.md once for the big-picture map of the field before starting the curriculum.
If you already know the role you want, skim career/genai-career-roles-universal.md so you can choose the best Part 2 track.

Every GenAI learner starts here, regardless of the target role.

#	Topic	Note	Est. Time	Difficulty
1	Python for AI	python-for-ai	3h	beginner
2	Linear Algebra	linear-algebra-for-ai	3h	beginner
3	Probability and Statistics	probability-and-statistics	2h	beginner
4	Neural Networks	neural-networks	3h	beginner
5	Deep Learning Fundamentals	deep-learning-fundamentals	2h	beginner
6	NLP Fundamentals	nlp-fundamentals	2h	beginner

#	Topic	Note	Est. Time	Difficulty
7	Transformers	transformers	4h	intermediate
8	Attention Mechanism	attention-mechanism	3h	intermediate
9	Tokenization	tokenization	2h	intermediate
10	Embeddings	embeddings	3h	intermediate
11	Modern Architectures	modern-architectures	4h	intermediate
12	Scaling Laws and Pretraining	scaling-laws-and-pretraining	4h	advanced

#	Topic	Note	Est. Time	Difficulty
13	LLMs Overview	llms-overview	3h	intermediate
14	Prompt Engineering	prompt-engineering	2h	beginner
15	Context Engineering	context-engineering	3h	intermediate
16	Function Calling and Structured Output	function-calling	3h	intermediate
17	RAG	rag	4h	intermediate
18	Fine-Tuning	fine-tuning	4h	intermediate
19	AI Agents	ai-agents	4h	intermediate
20	Evaluation and Benchmarks	evaluation	2h	intermediate

After Part 1, you have the foundation for any GenAI role.

Target roles: Full-Stack AI Engineer, AI Integration Engineer, AI Consultant, Prompt Engineer, AI Engineer

#	Note	Est. Time
A1	api-design-for-ai	2h
A2	graph-rag	3h
A3	vector-databases	3h
A4	multi-agent-architectures	3h
A5	agentic-protocols	4h
A6	conversational-ai	3h
A7	voice-ai	2h
A8	code-generation	3h
A9	ai-system-design	3h
A10	llmops	3h
A11	structured-outputs	3h
A12	ai-coding-agents	3h
A13	ai-product-management-fundamentals	2h

External skills: React or Next.js, Docker basics, cloud services, stakeholder communication

Target roles: GenAI Engineer, LLM Engineer, Agentic AI Engineer, RAG Engineer, AI DevTools Engineer

#	Note	Est. Time
B1	graph-rag	3h
B2	multi-agent-architectures	3h
B3	agent-evaluation	3h
B4	agent-memory	3h
B5	agentic-protocols	4h
B6	advanced-fine-tuning	4h
B7	rl-alignment	4h
B8	distillation-and-compression	3h
B9	synthetic-data	3h
B10	llm-evaluation-deep-dive	3h
B11	hallucination-detection	3h
B12	vector-databases	3h
B13	model-serving	3h
B14	monitoring-observability	3h
B15	cost-optimization	3h
B16	llmops	3h
B17	embedding-fine-tuning	3h
B18	long-context-engineering	3h
B19	retrieval-evaluation	3h
B20	guardrails-and-content-filtering	3h
B21	llm-routing-and-model-selection	3h
B22	document-parsing-and-extraction	2h
B23	data-flywheel-design	2h
B24	ai-ux-patterns	2h
B25	model-merging	3h

External skills: System design, Docker, Kubernetes, async programming

Target roles: ML Engineer, MLOps or LLMOps Engineer, AI Data Engineer, Data Scientist, Inference Optimization Engineer

#	Note	Est. Time
C1	docker-and-kubernetes	3h
C2	model-serving	3h
C3	monitoring-observability	3h
C4	cicd-for-ml	3h
C5	cloud-ml-services	3h
C6	ml-experiment-tracking	2h
C7	data-versioning-for-ml	2h
C8	classical-ml-for-genai	2h
C9	latency-and-throughput-engineering	3h
C10	distributed-systems-for-ai	3h
C11	distributed-inference-and-serving-architecture	3h
C12	inference-optimization	3h
C13	cost-optimization	3h

External skills: Docker, Kubernetes, cloud depth, DSA, PyTorch or TensorFlow

Target roles: Foundation Model Engineer, Research Scientist, Applied Scientist, AI Infrastructure Engineer

#	Note	Est. Time
D1	scaling-laws-and-pretraining	4h
D2	distributed-training	4h
D3	training-infrastructure	3h
D4	advanced-fine-tuning	4h
D5	rl-alignment	4h
D6	distillation-and-compression	3h
D7	interpretability	2h
D8	research-methodology-and-paper-reading	2h
D9	gpu-cuda-programming	4h
D10	reasoning-models	3h
D11	attention-deep-dive	4h
D12	state-space-models	3h
D13	model-merging	3h

External skills: C++, CUDA, distributed training, JAX, paper reading, reproducibility discipline

Target roles: AI Safety Engineer, AI Red Team Engineer, AI Ethics Lead, AI Data Governance Manager

#	Note	Est. Time
E1	ethics-safety-alignment	3h
E2	ai-regulation	2h
E3	adversarial-ml-and-ai-security	3h
E4	owasp-llm-top-10	2h
E5	hallucination-detection	3h
E6	evaluation-and-benchmarks	2h
E7	llm-evaluation-deep-dive	3h
E8	prompt-injection-deep-dive	3h
E9	mcp-security	3h

External skills: EU AI Act, NIST AI RMF, adversarial ML, secure AI delivery

Use these after Part 1 and your main track when you want deeper specialization, broader market awareness, or interview-focused capstones.

#	Note	Est. Time	Best For
S1	llm-landscape	3h	model selection, vendor comparisons, architecture trade-offs
S2	tools-overview	3h	tooling orientation before going deeper on infra or production

#	Note	Est. Time	Best For
S3	multimodal-ai	3h	builders moving beyond text-only systems
S4	computer-vision-fundamentals	3h	multimodal builders, CV-adjacent roles, document AI
S5	diffusion-models	4h	image generation, multimodal research, creative AI systems

#	Note	Est. Time	Best For
S6	continual-learning	2h	advanced adaptation, research-minded builders, lifelong agents

#	Note	Est. Time	Best For
S7	system-design-for-ai-interviews	2h	AI engineer, GenAI engineer, ML engineer, platform and research interview loops