LiminalML
ML / ResearchBackendFrontendSystem Design

The structured path to your
ML or SWE interview

Upload your resume. Get personalized STAR stories. Pick an ML or SWE topic and study it through a rigorous 6-stage session — from intuition to internals to production code — grounded in your own project experience.

157+
ML + SWE topics
6
Session stages
2 tracks:
2 tracks · 10 domains
AI-personalized to your resume

Two tracks. One study loop.

Pick what you're interviewing for — or flip between tracks mid-week.

ML / Research

71 topics · 5 domains

For Research Engineer, MLE, Research Scientist, and Applied Scientist loops. Stage 3 is full math — derivations with every term motivated.

Classical MLDeep LearningReinforcement LearningTraining & MLOps

Roles: Research Engineer · MLE · Research Scientist · Applied Scientist

Software Engineering

86 topics · 5 domains

For senior SWE, backend, frontend, and fullstack loops. Stage 3 swaps math for internals: data structures, complexity, tradeoffs, and failure modes.

FrontendBackendSystem DesignUI / UXCS Fundamentals

Roles: Frontend · Backend · Fullstack · System Design · UI/UX

Not a chatbot

Every session follows a rigid 6-stage structure. You can't skip stages, and the AI won't let you rush the derivation or the system design.

Grounded in your work

Your resume and STAR stories are injected into every session. The retrieval check asks about your specific implementation choices.

Interview-calibrated depth

Calibrated to someone who has read the papers or shipped the systems. Deeper, not shallower.

The 6-stage session structure

Every topic. Every session. Always in order. Never compressed.

Shown: ML track. The SWE track swaps Stage 3 “Math” for “Internals & Complexity”.

1Big Picture

2–3 sentences on what the concept solves and where it appears in real production systems. No jargon, no math — just the mental frame.

2Intuition + Visual

Core idea in plain language, then a structured diagram with tensor dimensions and data flow annotated. Mandatory for all DL architectures.

3The Math

Full step-by-step derivation with every term motivated. Not just what each symbol is, but what breaks if you remove it.

4Implementation

Production-quality PyTorch with type annotations, every non-obvious line commented, and an explicit test snippet at the end.

5Interview Questions

5 graded questions: conceptual, implementation, applied, systems-level (latency/memory/scale), and failure modes. Ordered from warm-up to hard.

6Retrieval Check

Conversational drill. The AI asks, waits for your answer, then tells you precisely what was right, wrong, or missing — no score, just a senior engineer interviewing you.

Between every stage

After each stage the AI pauses: “Ready to continue, or any questions before we move on?” Type “continue”to advance, or ask any clarifying question — the AI answers it fully, then resumes from where it left off. A quick-action button also appears so you don't have to type.

What Engineers Are Saying

JK

J.K.

ML Engineer, ex-Meta

The 6-stage structure is exactly what I needed. I stopped trying to memorize definitions and started actually deriving things. Passed my loop at a top lab.

AP

A.P.

Research Scientist, Google DeepMind

The resume-grounded retrieval check is genuinely useful. It asked me about my specific transformer project and I realized I couldn't explain my own work precisely enough.

DM

D.M.

Applied Scientist, Amazon

I've tried every ML prep resource. This is the only one that actually forces you to understand the math before moving on. The stage structure doesn't let you fake it.

Sessions grounded in your experience

Upload your resume PDF and the platform extracts your experience, projects, and skills, then generates 6–8 STAR stories mapped to common behavioral questions.

Every session injects your full profile into the system prompt. The Stage 6 retrieval check is grounded in your work — if you built an attention-based model or shipped a rate-limited service, the AI asks about your implementation choices, not a generic one.

Edit your STAR stories and extra context at any time on the Profile page. Changes apply to the next session immediately.

Resume upload

PDF → extracted text → STAR story generation

Profile injection

Full resume + stories injected into every session

Grounded retrieval

Stage 6 references your specific project work

Editable anytime

Update profile on Profile page, effective immediately

Session history

Every session is saved. Click any past session in the history sidebar to reload the full conversation and continue from where you left off. The sidebar shows topic, stage reached, and date.

Topic mastery is tracked locally — a colored dot in the topic browser shows topics you've studied and roughly how many sessions on each.

Revision cards

After Stage 2 or later, a “Revision card” button appears. Click it to generate a compact summary: core concept, key equations or tradeoffs, the one implementation detail that distinguishes strong candidates, and the two most likely interview questions.

Designed for rapid pre-interview review — the night before your loop.

71 topics across 5 domains

Classical models to LLM internals to RL theory — calibrated for RE / MLE / RS / AS loops.

Classical ML

16
  • Linear Regression
  • Logistic Regression
  • Support Vector Machines
  • K-Nearest Neighbours
  • K-Means Clustering
  • Naive Bayes
  • Decision Trees
  • Random Forests
  • Gradient Boosting / XGBoost
  • Gradient Descent (SGD, mini-batch, momentum, Adam)
  • Bias-Variance Tradeoff
  • Regularization (L1, L2, Dropout)
  • Cross-Validation
  • Evaluation Metrics (precision, recall, F1, AUC-ROC)
  • Feature Engineering
  • Dimensionality Reduction (PCA, t-SNE)

Deep Learning

26

Foundations

  • Backpropagation and Computational Graphs
  • Activation Functions
  • Loss Functions
  • Optimizers (Adam, AdamW, LR schedules)
  • Batch Normalization and Layer Normalization
  • Weight Initialization

Architectures

  • Feedforward Networks
  • CNNs
  • RNNs, LSTMs, GRUs
  • Transformers (encoder, decoder, encoder-decoder)
  • Attention Mechanisms (scaled dot-product, multi-head, cross-attention, FlashAttention)
  • Positional Encoding
  • KV Cache

Language and LLMs

  • Tokenization and BPE
  • Embeddings (word2vec, learned)
  • Pretraining vs Fine-tuning
  • LoRA and PEFT
  • RAG (Retrieval Augmented Generation)
  • Tool Calling
  • Temperature Sampling, Top-k, Top-p, Beam Search
  • RLHF (reward modeling, PPO fine-tuning loop, preference data)
  • Inference Optimization

Multi-task and Transfer Learning

  • Transfer Learning
  • Multi-task Learning
  • Negative Transfer
  • Task Sampling Strategies

Reinforcement Learning

14

Foundations

  • Markov Decision Processes
  • Bellman Equations
  • Value Functions
  • Policy-based vs Value-based vs Model-based RL

Tabular Methods

  • Dynamic Programming
  • Monte Carlo Control
  • Temporal Difference Learning
  • Q-Learning
  • SARSA

Deep RL

  • DQN
  • Policy Gradients (REINFORCE)
  • Actor-Critic Methods
  • PPO (Proximal Policy Optimization)
  • DPO (Direct Preference Optimization)

Training Engineering

8
  • Gradient Accumulation
  • Gradient Checkpointing
  • Mixed Precision Training (fp32, fp16, bfloat16)
  • CUDA Setup and Device Management
  • Distributed Training and DDP
  • Training Config and YAML Structure
  • Multi-Seed Evaluation and Reproducibility
  • Experiment Tracking with MLflow

Systems and MLOps

7
  • Model Serving and Deployment
  • Embeddings at Scale
  • Vector Databases
  • Batch vs Online Inference
  • Latency vs Throughput Tradeoffs
  • CI/CD for ML
  • Experiment Tracking

Pricing

Simple, honest pricing

Start free. Upgrade when the monthly cap gets in the way. Cancel any time.

Free
$0/ month

Generous monthly free tier. Ideal for casual prep or trying the platform.

  • 20 sessions per month
  • All 130+ ML and SWE topics
  • Full 6-stage session format
  • Resume-personalized STAR stories
  • Session history + reload
  • Revision cards
Pro
7-day free trial
$9/ month

For candidates running daily sessions in the weeks leading up to interview loops.

  • Unlimited sessions
  • Everything in Free
  • Priority access to new topics
  • 7-day free trial
  • Cancel anytime

No charge for 7 days. Cancel anytime in your profile.

Start preparing today

Upload your resume and run your first session in under 5 minutes.

Get started