ML Engineer

Interview-ready on fine-tuning, alignment, and LLM evaluation

ML engineer interviews now include deep LLM questions: "When would you fine-tune vs. use RAG?", "Explain LoRA," "How do you evaluate a fine-tuned model?" This path covers transformer math, PEFT techniques, training data curation, alignment methods (RLHF, DPO), and evaluation frameworks — everything you need for ML engineering interviews in the LLM era.

ML EngineersData Scientists moving into LLMsResearch EngineersApplied ScientistsNLP Engineers
1Free Modules
7Premium Modules
8Roadmap Steps

Your Learning Path

A step-by-step roadmap from foundations to mastery. Follow this sequence for the most effective learning experience.

Map out the ML engineer LLM skills landscape and career trajectory
Deep dive into transformer architecture and modern LLM variants
Develop a decision framework: when to fine-tune vs. prompt vs. RAG
Master parameter-efficient fine-tuning methods (LoRA, QLoRA)
Learn to curate and build high-quality training datasets
Run fine-tuning jobs on cloud GPUs with proper experiment tracking
Implement alignment techniques (RLHF, DPO) for instruction following
Build comprehensive evaluation pipelines for model quality

Modules

8

1 free module to get you started, plus 7 premium deep-dives.

1Free

ML Engineer Roadmap

The complete learning path for ML engineers working with LLMs: from transformer fundamentals to production fine-tuning and alignment. Understand the difference between GenAI engineering and ML engineering roles.

15 minStart
2Premium

LLM Architecture Deep Dive

Transformer architecture at the mathematical level: self-attention equations, multi-head attention, positional encodings (RoPE, ALiBi), layer normalization, feed-forward networks, and how modern LLMs (GPT, Llama, Claude) differ architecturally.

60 minUpgrade to access
3Premium

The Fine-Tuning Decision

When to fine-tune vs. use prompting vs. RAG. Cost-benefit analysis frameworks, data requirements estimation, compute budgeting, and a decision tree for choosing the right approach for your use case.

30 minUpgrade to access
4Premium

Parameter-Efficient Fine-Tuning (PEFT)

Deep dive into LoRA, QLoRA, prefix tuning, adapters, and IA3. Understand the math behind low-rank adaptation, how to choose rank and alpha hyperparameters, and when each PEFT method shines.

60 minUpgrade to access
5Premium

Training Data for Fine-Tuning

Building high-quality fine-tuning datasets: data collection strategies, annotation guidelines, quality filtering, synthetic data generation, data formatting (Alpaca, ShareGPT, chat templates), and dataset evaluation.

45 minUpgrade to access
6Premium

Running Fine-Tuning Jobs

Hands-on fine-tuning execution: HuggingFace Transformers + TRL, Axolotl, cloud GPU provisioning (Lambda Labs, RunPod, AWS), hyperparameter tuning, distributed training basics, and experiment tracking with W&B.

60 minUpgrade to access
7Premium

Alignment: RLHF, DPO, ORPO

How models learn to follow instructions and be helpful. Reward model training, PPO for RLHF, Direct Preference Optimization (DPO), Odds Ratio Preference Optimization (ORPO), constitutional AI, and building preference datasets.

60 minUpgrade to access
8Premium

Model Evaluation

Comprehensive LLM evaluation: automated benchmarks (MMLU, HumanEval, MT-Bench), human evaluation protocols, task-specific metrics, LLM-as-judge, regression testing, and building evaluation pipelines for fine-tuned models.

45 minUpgrade to access

Start Free — No Account Required

These foundational resources are free for everyone. Build your AI literacy before diving into persona-specific modules.

Prompt 101GenAI GlossaryBest Free ResourcesCertificationsUse Cases

Unlock All 7 Premium Modules

Get full access to every ML Engineer module — plus all other GenAI personas, DSA content, and System Design content with a single subscription.

View Pricing