Paper of the week
New paper
#1
VFlash: Fast, Accurate Visual Planning via Flash Primitive Decoding
· NVIDIA, 2026
Why it matters
u_
Unrolled
System Design
55 Cases
DSA for AI/ML
New
Curriculum
Practice
Papers
Pricing
More
Library
Every lesson,
by topic.
239 written lessons, each with runnable code and an exercise. Use it to look things up. To learn in order, follow Build Atlas.
New here? Learn in order instead.
Build Atlas takes these lessons from zero to a GPT you train, fine-tune and ground in documents.
Search lessons
CS & Systems Foundations
30 lessons
Values, Names and Types
What a Variable Really Is
Free
15m
Control Flow
Decisions, Loops and Loops That Know When to Stop
Free
15m
Functions
Arguments, Scope and the Call Stack
Free
15m
Lists and Tuples
Sequences, Copies and What They Cost
Free
15m
Dicts and Sets
Hashing and O(1) Lookups
Free
15m
Strings and Unicode
Text Is Harder Than It Looks
Free
15m
How Python Really Runs Your Code
Atlas starts life as a Python program that stores lessons and answers questions.
Free
20m
The GIL
Threads, Processes and Which One to Use
Free
15m
C++ and Memory
What Python Hides From You
Free
20m
Data Structures That Power AI Systems
Thousands of students use Atlas at once. Each request must find that student's history (lookup), wait its turn fairly (scheduling), and fetch the best-matching lessons (ranking).
Free
20m
Comprehensions, Iterators and Generators
Data Too Big for Memory
Free
15m
Closures, Decorators and Context Managers
Reusable Behaviour
Free
15m
Classes, Dataclasses and Type Hints
Modelling Your Domain
Free
15m
Errors, Logging, Tests and Packaging
Code You Can Trust
Free
15m
NumPy and Pandas
Vectorized Python for Data and ML
Free
20m
Profiling
Find the Slow Line Before You Fix Anything
Free
15m
Async Python
Thousands of LLM Calls Without Thousands of Threads
Free
15m
Interview Patterns and Capstone
An Async, Rate-Limited LLM Batch Client
Free
15m
Operating Systems
Sharing One Machine Between Many Programs
Free
10m
Networking
How a Question Travels to the Model and Back
Free
10m
Databases
Remembering Things Safely and Finding Them Fast
Free
20m
SQL Basics
Tables, SELECT, WHERE, ORDER BY and the Truth About NULL
Free
15m
SQL Joins
Inner, Left, Semi and Anti Joins, and Subqueries vs Joins
Free
15m
SQL Aggregation
GROUP BY, HAVING, Dates, Buckets and Pivots
Free
15m
SQL Window Functions
Ranking, Running Totals, Top-N and Gaps and Islands
Free
15m
SQL CTEs and Recursive Queries
Readable Pipelines, Hierarchies and Funnels
Free
10m
SQL Under the Hood
Query Plans, Indexes, Columnar Engines and Transactions
Free
20m
SQL for ML
Point-in-Time Training Sets, A/B Test Analysis and Vector Search in Postgres
Free
15m
SQL Interview Patterns and Capstone
A Product Analytics Warehouse
Free
15m
Bridge: From Storing Text to Understanding It
Atlas can store and serve everything — but understands nothing it stores.
Free
10m
Math, Stats & Optimization
12 lessons
Reading Maths Notation
Functions, Sets, Sums, Products and argmax
Free
15m
Linear Algebra
The Language of Embeddings and Attention
Free
20m
Linear Systems, Rank, Inverses and Eigenvectors
Three questions recur across ML: *can this system be solved, and is the answer unique?* (rank and inverses: least squares, redundant features), *what does this matrix do when applied again and again?
Free
15m
SVD and PCA
Every Matrix Is a Rotation, a Stretch and a Rotation
15m
Calculus
How a Model Knows Which Way to Improve
Free
15m
Jacobians, Matrix Calculus and Taylor Series
Gradients of Whole Layers and Local Approximations
15m
Probability & Statistics
Reasoning When You're Not Sure
Free
15m
Joint Distributions, the Gaussian, Sampling and the Central Limit Theorem
Real data has many variables at once, so we need joint distributions: how variables vary together (covariance, correlation), what one tells us about another (conditionals), and when they're unrelated (independence).
15m
Optimization
Turning Gradients Into Learning
Free
15m
Convexity and Lagrange Multipliers
When Local Means Global, and Optimizing Under Constraints
15m
Information Theory
Measuring Surprise, the Loss Every LLM Minimizes
Free
15m
Bridge: From Formulas We Tune to Models That Learn
Atlas can represent meaning as vectors and minimize a loss — but every formula so far was designed by us.
10m
Classical ML
11 lessons
Supervised Learning
Learning From Labelled Examples
10m
Three Classic Classifiers
k-Nearest Neighbours, Naive Bayes and Support Vector Machines
10m
Boosting
From AdaBoost to Gradient Boosting, XGBoost, LightGBM, CatBoost and Stacking
15m
Unsupervised Learning
Finding Structure Without Labels
10m
Gaussian Mixtures and EM, t-SNE and UMAP
Soft Clusters and Honest Pictures of High-Dimensional Data
10m
ML Problem Families
Choosing the Right Shape for the Problem
Free
15m
Feature Engineering
Turning Raw Data Into Model Inputs
Free
15m
ML Evaluation
Measuring Models Without Fooling Yourself
Free
15m
Imbalanced Data, Hyperparameter Search and SHAP
Making Tabular Models Work and Explaining Them
10m
Classical ML Capstone
Churn Prediction, Price Prediction, a Recommender, and the ML Interview
10m
Bridge: When Hand-Made Features Stop Scaling
Atlas's classic models work on tidy tables — but they can't read a question's text, an image or a sequence without someone hand-crafting features.
10m
Deep Learning
6 lessons
Neural Network Fundamentals
Layers That Learn Their Own Features
Free
15m
Autograd From Scratch, and Debugging Training
Gradient Checks, Loss Curves and the Usual Bugs
15m
Major Architectures
Matching the Network to the Shape of the Data
15m
CNNs From Scratch and ResNets
Convolution, Pooling, Residuals and Modern CNN Design
15m
PyTorch Engineering
From Tensors to Fast, Correct Training Loops
20m
Bridge: From Learning Numbers to Learning Language
Atlas can train deep networks efficiently — but its input is still numbers we prepared. Students type text.
10m
LLM Architecture
7 lessons
NLP Foundations
From Text to Tokens to Predictions
15m
The Transformer I
Attention, Positions and Heads from First Principles
Free
25m
The Transformer II
The Block, the Output, the Arithmetic and Production Practice
Free
30m
LLM Architecture Families
Trade-offs, Not Model Names
15m
LLM Pretraining
Data, Compute and Scaling Laws
15m
Build an LLM from Scratch
Every idea in this domain has been shown in pieces. This lesson puts them together into one script that trains a working GPT: data → tokenizer → model → training → checkpoint → sampling → evaluation.
25m
Bridge: From a Model That Writes to a Model That Helps
Atlas now has a language model that writes fluent text. Ask it a question, though, and it carries on writing instead of answering.
10m
Post-training & Alignment
10 lessons
Supervised Fine-Tuning
Teaching a Base Model to Answer
25m
Parameter-Efficient Fine-Tuning
LoRA, QLoRA and Adapters
20m
Preference Optimization
Reward Models, RLHF and DPO
20m
Reinforcement Learning I
Bandits, Exploration, MDPs and the Bellman Equation
Free
10m
Reinforcement Learning II
Monte Carlo, TD Learning, Q-Learning and DQN
Free
15m
Reinforcement Learning III
Policy Gradients, Baselines, Actor-Critic, GAE and PPO From Scratch
15m
Reinforcement Learning for LLMs
Policy Gradients, PPO and GRPO
20m
RL in Practice
Environments, Verifiers, Reward Hacking, Reading RL Curves, and a GRPO Capstone
15m
Synthetic Data
Distillation, Self-Instruct and Rejection Sampling
20m
Bridge: From Changing the Weights to Changing the Input
Atlas is now an assistant: fine-tuned, preference-tuned and trained with verifiable rewards. But every new feature request can't mean another training run.
10m
Prompt & Context Design
5 lessons
Prompt Fundamentals
Instructions a Model Can Actually Follow
Free
20m
Advanced Prompting
Reasoning, Sampling Many Answers and Constrained Output
Free
15m
Context Engineering
Deciding What the Model Sees
Free
20m
Prompt Evaluation
Knowing Whether a Change Helped
20m
Bridge: From What You Put in the Prompt to Finding It
Atlas can now write, structure, budget and test its prompts. But the best prompt is useless if the passage that answers the student's question never makes it into the context.
10m
RAG & Knowledge Systems
8 lessons
Retrieval I
The Pipeline, the Inverted Index and BM25 from First Principles
25m
Retrieval II
Analyzers, Speed, Sharding, Permissions and Debugging
35m
Embeddings
Turning Meaning into Geometry
20m
Document Ingestion
From Messy Files to Clean, Retrievable Chunks
20m
Vector Databases
Finding Nearest Neighbours Among Millions
20m
Advanced RAG
Hybrid Search, Reranking, Context and Citations
20m
RAG Evaluation
Measuring Retrieval and Answers Separately
20m
Bridge: From Answering Questions to Taking Actions
Atlas can now find the right passage and answer from it, with citations and measured quality. But students don't only ask questions.
10m
AI Agents & Tool Use
11 lessons
Agent Fundamentals
The Loop That Lets a Model Act
20m
Agent Architectures
Routers, Planners, Critics and Teams
20m
Tool & Protocol Engineering
Tools Agents Can Use Safely
20m
Agent Memory
What to Keep, Where, and How to Forget
20m
Advanced Agent Engineering
Durable, Parallel, Budgeted, Observable
20m
Reasoning & Test-Time Compute
Thinking Longer Instead of Training Bigger
20m
The Model Context Protocol
One Plug for Every Tool
20m
Reasoning Failures
How Agents Go Wrong, and How to Catch It
20m
Capstone: A Production Support Assistant
and Choosing Between Prompting, RAG and Fine-Tuning
15m
Capstone: A Research Agent with Tools and Memory
and Agentic RAG That Decides When to Search
10m
Bridge: From an Agent That Works to Proof That It Works
Atlas can now act: it reads progress, writes quizzes, checks code and handles refunds, with guards, memory and traces. The team wants to switch to a cheaper model and a new prompt next week.
10m
AI Evaluation & Evals
5 lessons
Evaluation Foundations
Measuring Quality You Can Trust
15m
LLM Evaluation
Benchmarks, Contamination and Judges
20m
Agent Evaluation
Did It Do the Right Thing, the Right Way, Every Time?
15m
The Production Evaluation Pipeline
Gates, A/B Tests and Live Monitoring
20m
Bridge: From Measuring Quality to Running a Machine Learning System
Atlas has trustworthy evals, release gates and live monitoring.
10m
Production ML Systems
8 lessons
The ML Lifecycle
From a Notebook to a System That Keeps Working
15m
ML Pipelines
Reproducible, Incremental, Validated
15m
ML Platform Tools
What MLflow, DVC, Docker, Ray and Friends Actually Do
15m
Feature Infrastructure
Point-in-Time Correct, Fresh, Consistent
15m
Model Deployment
Batch, Online, Async, and Shipping Without Fear
15m
Model Monitoring
Drift, Delayed Labels and Alerts People Trust
20m
Capstone: An End-to-End ML Platform
From Raw Events to Monitored, Self-Refreshing Models
10m
Bridge: From Serving Models to Serving LLMs
Atlas now has pipelines, a feature store, safe rollouts and monitoring. Then the team self-hosts an open 8B model for the tutor to cut API costs.
10m
Inference Serving & vLLM
17 lessons
Inference Fundamentals
Prefill, Decode and the Roofline
35m
KV Cache I
What It Stores, Exact Sizing, and Paging
35m
KV Cache II
Prefix Sharing, Preemption, Quantization, Capacity and Operations
35m
KV Cache III
Long Contexts, Attention Sinks and Eviction
20m
Batching & Scheduling
Continuous Batching, Chunked Prefill and Fairness
30m
Quantization
Fewer Bits per Number, Measured Honestly
30m
Advanced Inference Optimization
Speculation, Fused Attention, Parallelism and MoE
25m
Inference Frameworks
Inside vLLM, SGLang, TensorRT-LLM and Friends
30m
Measurement & Capacity
Benchmarks That Don't Lie, and Breaking One GPU on Purpose
30m
Observability for LLM Serving
Golden Signals, Prometheus and SLOs
20m
Disaggregated Serving
Splitting Prefill and Decode Across Machines
25m
Tiered KV Cache
HBM, CPU Memory and SSD for Long Contexts and Reuse
15m
Inference Economics
What a Token Costs, and Which Levers Move It
25m
Autoscaling LLM Servers
Queue-Depth Signals, KEDA and Cold Starts
20m
Routing and Token Budgets
Model Cascades, Rate Limits and Load Shedding
20m
The Inference Reading List
The Papers Behind Every Serving Engine
15m
Bridge: From Serving Systems to the Silicon Underneath
Atlas's serving stack now batches, pages, quantizes, speculates and caches across tiers.
10m
GPU & CUDA Programming
8 lessons
GPU Architecture
Thousands of Simple Cores and a Memory System to Feed Them
15m
The CUDA Programming Model
Grids, Blocks, Threads and Streams
15m
The GPU Memory Hierarchy
Coalescing, Shared Memory and Bank Conflicts
15m
Parallel Computation
Reductions, Scans, Atomics and Tiled Matmul
20m
GPU Performance Engineering
Occupancy, Profiling and the Roofline in Practice
15m
CUDA Libraries & Tools
cuBLAS, cuDNN, CUTLASS, NCCL and When Not to Write a Kernel
15m
Triton Programming
Fast Fused Kernels in Python
20m
Bridge: From One GPU to Many Machines
Atlas's kernels are near the roofline and one GPU is used well. But Atlas now runs on 60 GPU servers, a queue of nightly jobs, three databases and a vector store, in two regions.
10m
Distributed Systems
6 lessons
Distributed Systems Fundamentals
Failure, Time, Consistency and Consensus
20m
Messaging & Event Processing
Queues, Logs and Exactly-Once Effects
15m
Distributed Storage
Replication, Partitioning and Object Stores
15m
Distributed Compute
MapReduce, Spark, Ray and the Straggler Problem
15m
Reliability Engineering
SLOs, Redundancy, Overload and Incidents
15m
Bridge: From Distributed Services to One Training Job on Thousands of GPUs
Atlas's services now replicate, partition, queue, retry and degrade gracefully. The research team wants to train Atlas's own 70B-parameter tutor model on 2 trillion tokens.
10m
Distributed Training
6 lessons
Parallelism Strategies
Data, Tensor, Pipeline, Sequence and Expert
15m
Collective Communication
All-Reduce, All-Gather and the Networks Underneath
15m
Training Memory Optimization
ZeRO, FSDP, Checkpointing and Mixed Precision
15m
Training Infrastructure
Keeping 4,000 GPUs Productive for Weeks
15m
Distributed Training Frameworks
DDP, FSDP2, DeepSpeed, Megatron and JAX
15m
Bridge: From Training at Scale to the Data That Feeds It
Atlas's training stack can now keep 2,048 GPUs busy for weeks. The research lead asks the obvious next question: busy on what?
10m
AI Data Engineering
4 lessons
Data Ingestion & Storage
Lakes, Warehouses, Columnar Files and Table Formats
15m
Data Processing & Quality
Cleaning, Deduplication, Filtering and PII at Scale
20m
Datasets for LLMs
Mixtures, Repetition, Instruction Data and Versioning
15m
Bridge: From Data and Models to the Platform They Run On
Atlas now has versioned datasets, training jobs, inference servers, feature pipelines, vector stores and evaluation jobs.
10m
Cloud & AI Platform
5 lessons
Cloud Fundamentals
Compute, Storage, Networks, Identity and Cost for AI
15m
Kubernetes for ML
Reconciliation, GPU Scheduling and Batch Jobs
15m
Serving Platforms
Autoscaling, Cold Starts and the LLM Gateway
15m
Infrastructure as Code
Plans, State, Drift, GitOps and Policy
15m
Bridge: From a Platform for Text to Inputs That Aren't Text
Atlas's platform now runs reproducibly in two regions: cost-aware purchasing, a shared GPU fleet with gang scheduling, autoscaled serving behind a gateway, and everything in code.
10m
Multimodal Foundation Models
8 lessons
Computer Vision
From Pixels and Filters to Detection and Segmentation
20m
Vision in Production
A Defect-Detection Project with YOLO, and Serving Models on the Edge
10m
Audio & Speech
Spectrograms, ASR, TTS and Real-Time Voice
15m
Video Intelligence
Sampling, Temporal Models and Video-Language Systems
15m
Spatial AI & Robotics
3D Perception, Control Loops and Vision-Language-Action Models
15m
Vision-Language Models
CLIP, Multimodal LLMs and Document Understanding
15m
Generative Image Models
Autoencoders, VAEs and Diffusion
15m
Bridge: From More Inputs to More Ways In
Atlas now reads photos, screenshots and scanned exams, listens to voice notes, indexes lecture videos, drives a lab robot through a safety watchdog, and generates diagrams.
10m
AI Security & Safety
5 lessons
AI Security Fundamentals
Threat Models, Poisoning, Supply Chain and Privacy Attacks
15m
LLM Attacks
Jailbreaks, Prompt Injection, Exfiltration and Red-Teaming
15m
Agent Security
Least Privilege, Trust Boundaries, Sandboxing and Human Confirmation
15m
Responsible AI
Fairness, Privacy, Transparency and Governance in Engineering Terms
15m
Bridge: From Defending the System to Advancing It
Atlas is now threat-modelled, red-teamed, permission-bounded and governed.
10m
Research Engineering
6 lessons
Reading Papers
Three Passes, Checking Claims and Knowing What to Trust
15m
Research Workflow
Hypotheses, Ablations, Seeds, Significance and Scaling Studies
15m
Reproducing Papers
From PDF to a Verified Number, and Writing It Up
10m
Interpretability
Probes, Activation Patching, Superposition and Sparse Autoencoders
15m
World Models and JEPA
Predicting in Latent Space and Planning with Learned Dynamics
15m
Bridge: From Doing the Work to Multiplying It
Atlas now has every technical layer: data, models, serving, agents, evaluation, security and a research practice that can read, test and reproduce new ideas.
10m
Staff Engineering & Leadership
6 lessons
Technical Leadership
Strategy, Choosing Problems and Running Multi-Team Work
10m
System Design Thinking
Boundaries, Interfaces, Failure Domains and Evolution
10m
Engineering Judgment
Decisions Under Uncertainty, Reversibility, Build vs Buy and Technical Debt
15m
Technical Writing
Design Docs, Decision Records, Postmortems and Strategy Docs
15m
Cross-Team Impact
Influence Without Authority, Platforms and Migrations
10m
Bridge: From Leading the Work to Designing Whole Systems
Atlas's growth plan says ten million students in two years, plus licensing its recommendation and fraud models to partner platforms.
10m
AI System Design Bank
55 lessons
The System Design Method
From a Vague Prompt to a Defensible ML System at Staff Level
Free
15m
Capacity Estimation and Sizing
Back-of-the-Envelope Numbers for ML Systems at Staff Level
20m
Feature Stores and Training–Serving Parity
Point-in-Time Data, Freshness Tiers and Skew You Can Measure at Staff Level
20m
Case Study: Fraud and Abuse Detection
Cost-Based Decisions, Real-Time Features and Delayed Labels at Staff Level
25m
Case Study: Spam Filtering
Tiered Actions, Cascades and Campaigns at Staff Level
25m
Case Study: Churn Prediction
Clean Labels, Uplift and Measured Interventions at Staff Level
25m
Case Study: Surge Pricing
Forecasts, a Stable Price Controller and Marketplace Experiments at Staff Level
25m
Case Study: ETA Prediction and Routing
Live Traffic, Graph-Plus-ML Correction and Honest Promises at Staff Level
25m
Case Study: Sensor Anomaly Detection and Predictive Maintenance
Edge Features, Normal-Behaviour Models and Alert Budgets at Staff Level
25m
Case Study: Credit Underwriting
Delayed Labels, Explainable Models and Fair, Economic Decisions at Staff Level
25m
Case Study: Patient Readmission Risk
Leak-Free Labels, Capacity-Aware Targeting and Proving the Care Helped at Staff Level
25m
Case Study: Energy Load Forecasting
As-Issued Weather, Per-Horizon Models and Honest Intervals at Staff Level
25m
Case Study: Ad Click-Through Rate Prediction
Calibration, Freshness and the Auction at Staff Level
25m
Case Study: Recommendations and Search on One Platform
Shared Retrieval, Graded Labels and Bias Correction at Staff Level
20m
Case Study: Short-Video Recommendation Feed
Fresh Retrieval, Multi-Task Ranking and Long-Term Guardrails at Staff Level
25m
Case Study: E-commerce Search and Substitution
Availability-Aware Retrieval, Learned Ranking and Constraint-Safe Replacements at Staff Level
25m
Case Study: Music Discovery Playlists
Taste Clusters, Cold-Start Audio and Whole-Playlist Design at Staff Level
25m
Case Study: Job and Candidate Matching
Mutual Fit, Congestion Control and Fair Hiring Signals at Staff Level
25m
Case Study: Friend Suggestions on a Social Graph
Pruned Candidates, Meaningful-Connection Labels and Privacy by Design at Staff Level
20m
Case Study: News Feed Ranking
Hybrid Fan-Out, Value Models and a Separate Integrity Layer at Staff Level
20m
Case Study: Vacation Rental Search
Availability as a Hard Filter, Booking-Aware Ranking and a Fair Marketplace at Staff Level
25m
Case Study: Food Delivery Dispatch
Predicting Preparation, Batch Matching and Fair Offers to Couriers at Staff Level
25m
Case Study: Search Autocomplete
Precomputed Prefixes, Fresh and Personal Ranking, and Suggestions That Are Safe to Say at Staff Level
20m
Case Study: Visual Search
Routing, Embeddings Across the Photo Gap, Compressed Indexes and Privacy Rules at Staff Level
20m
Case Study: Real-Time Ad Bidding
Valuation in Milliseconds, First-Price Shading and Budgets That Never Overspend at Staff Level
25m
Case Study: Question Deduplication
Intent-Level Matching, Tiered Actions and Humans in Charge at Staff Level
20m
Case Study: Dating App Matchmaking
Reciprocal Ranking, Attention Balancing and Safety First at Staff Level
20m
Case Study: App Store Search
Navigational Intent, Retention-Based Labels and Integrity Against Gaming at Staff Level
20m
Case Study: An LLM Assistant Platform
Token Planning, Routing, Grounding and Safe Actions at Staff Level
25m
Case Study: Multi-Tenant Enterprise RAG
Permissions, Freshness and Cost at Staff Level
45m
Case Study: Inline Code Completion Engine
Latency, Context and GPU Cost at Staff Level
35m
Case Study: Customer Support Agent
Actions, Policy and Hand-off at Staff Level
30m
Case Study: Legal Document Analysis
Extraction with Evidence, Recall and Confidentiality at Staff Level
30m
Case Study: Clinical Q&A with Citations
Grounding, Safety Tiers and Patient Privacy at Staff Level
30m
Case Study: Financial Filings Summarizer
Exact Numbers, Change Detection and Minutes-After-Release at Staff Level
30m
Case Study: Multimodal Voice Assistant
Turn-Taking, Device–Cloud Split and Safe Actions at Staff Level
30m
Case Study: Autonomous Software Engineering Agent
Sandboxes, Verification and Cost per Task at Staff Level
30m
Case Study: Text-to-SQL Analyst
Semantic Layers, Correctness and Warehouse Cost at Staff Level
30m
Case Study: Long-form Document Synthesis
Planning, Consistency and Claim-level Citations at Staff Level
30m
Case Study: Multi-agent Research Engine
Orchestration, Token Budgets and Web Safety at Staff Level
30m
Case Study: Web Automation Agent
Perception, Irreversible Actions and Injection Defence at Staff Level
30m
Case Study: Email Assistant with Drafts
Triage, Voice and Inbound Injection at Staff Level
25m
Case Study: Meeting Copilot with Actions
Transcription, Attribution and Consent at Staff Level
30m
Case Study: High-throughput LLM Gateway
Token Quotas, Failover and Overhead at Staff Level
30m
Case Study: Multi-model Semantic Router
Cost–Quality Routing, Shadow Evaluation and Drift at Staff Level
25m
Case Study: Multi-tenant LoRA Serving
Adapter Batching, Memory Tiers and Isolation at Staff Level
30m
Case Study: Prefix Caching KV Store
Block Keys, Prefix-Aware Routing and Memory Tiers at Staff Level
35m
Case Study: Speculative Decoding Engine
Acceptance Rates, Lossless Verification and Load-Aware Control at Staff Level
30m
Case Study: LLM Guardrail and Safety Gateway
Check Cascades, Streaming Moderation and Failure Policy at Staff Level
30m
Case Study: Real-time Vector Database at Scale
Segments, Filtered Search and Deletion Guarantees at Staff Level
30m
Case Study: LLMOps Evaluation Gates
Release Bundles, Statistical Gates and Canaries at Staff Level
30m
Case Study: Synthetic Data and Distillation
Prompt Coverage, Quality Filters and Escalation at Staff Level
30m
Case Study: Privacy-preserving LLM Gateway
Data Classes, Routing, Pseudonymisation and Retention at Staff Level
30m
Case Study: Geo-distributed LLM Serving
Residency Islands, Shared GPU Pools and Regional Failover at Staff Level
30m
Capstone: Atlas at 10 Million Students
One Design to Fund, Run and Grow at Staff Level
25m