Mentoring · LLM & AI Engineer
LLM and AI engineering mentoring
Learn to build LLM systems that can be measured and shipped: from the transformer you wrote yourself to RAG with verifiable sources, evaluation before release and local serving.
Who this is for
- ML engineers and data scientists moving into LLM and AI engineering.
- Backend developers who build products on top of language models.
- Engineers who want to go beyond prompts: fine-tuning, evaluation and serving.
What you will build
The LLM & AI Engineering track has 8 sprints and builds on the Deep Learning stage, where you write a small GPT and pretrain a language model yourself.
- From the transformer to controllable generation.
- SFT and LoRA: an adapter with a quality check.
- Structured output and function calling.
- RAG: context, retrieval and verifiable sources.
- LLM evaluation and regression tests before release.
- Local serving: GGUF, quantisation and vLLM.
- An agent workflow with limits and safe tools.
- Observability, failures and reproducible releases.
How mentoring works
- A free intro call; we check whether you need Deep Learning first.
- Weekly 1:1 calls and personal code review of every sprint.
- Portfolio projects and mock interviews for AI engineering roles.
Questions
Do I need deep learning experience?
It helps. If you have none, we go through the Deep Learning stage first, where you implement a transformer and train a small model.
Is this about prompt engineering?
Partly, but most of the work is engineering: fine-tuning, retrieval, evaluation, serving and monitoring.
Which models do we use?
Open models you can run on free GPUs in Colab or Kaggle, so every result is reproducible.
Read next
LLM engineer roadmap: how to become an AI engineerAn LLM engineer roadmap: transformers, fine-tuning with LoRA, RAG, tool calling, agents, evaluation, serving and cost — with portfolio projects that prove the skills.Machine learning engineer roadmap: from Python to productionA machine learning engineer roadmap: Python and data, maths, classic ML, deep learning, a specialisation and MLOps, with portfolio projects and realistic timelines.Machine learning interview preparation: questions and a planMachine learning interview preparation: ML theory questions with short answers, coding, ML system design walkthroughs and a six-week plan for ML engineer interviews.
Mentoring directions
Quant Research1:1 quant research mentoring from a working quantitative researcher: statistics, alpha research, backtesting, portfolio and risk, and quant interviews.ML Engineer1:1 machine learning mentoring: Python, maths, classic ML on Kaggle, deep learning, a specialisation and ML system design, with personal code review.GPU & ML Performance1:1 mentoring in GPU programming and ML performance: CUDA and Triton kernels, fused softmax, tiled matmul, autotuning and custom PyTorch operators.Quant Developer1:1 quant developer mentoring in modern C++: performance, concurrency, networking, a FIX gateway, order book matching, market data replay and QuantLib.Interview PrepPrepare for ML engineer and quant research interviews: mock interviews, CV review and a trainer with 2,000 questions and spaced repetition. In English or Russian.