# insideout > insideout is a learning site for engineers. Lessons that show how AI and quant systems work and how systems are kept secure, with every step of every example written out. Around them: daily practice, interview prep round by round, and projects you build yourself. Name: insideout, one lowercase word, at insideout.sh. People also write it as inside out or InsideOut, and search for it as "insideout course", "inside out AI" or "inside out quant". It is not connected to the Pixar film Inside Out or to the Inside-Out Prison Exchange Program. Elsewhere: X https://x.com/insideoutdotsh, Instagram https://www.instagram.com/insideoutdotsh/. Who it's for: Engineers who want to understand the tools they use: developers moving into AI engineering, anyone heading into quant research, trading or development, and anyone starting or growing a career in security. Free: The first three lessons of every AI and quant module, with no account. A free account adds Practice and Interview, and saves your progress on any device. Projects and Cybersecurity come with Pro. Pro: Every lesson in every module, every interview round, and unlimited practice: more sets, any topic on its own, and timed runs. You may quote and summarise any public page with a link to it. Pro lessons show their opening and first part to everyone; the rest is for Pro members and is not in this file. ## Start here - [About insideout](https://insideout.sh/about): insideout is an online learning site for AI engineering, quant engineering and cybersecurity: lessons with every step shown, daily practice, interview prep and projects. - [Learn](https://insideout.sh/curriculum): Courses in AI engineering, quant engineering and cybersecurity from insideout: short interactive lessons with every step shown. The first three lessons of every AI and quant module are free. - [Tracks](https://insideout.sh/tracks): Three tracks through insideout's lessons: AI engineering, quant engineering and cybersecurity. Every route is free to see, and you can skip what you already know. - [Glossary](https://insideout.sh/glossary): The words of AI engineering, quant engineering and cybersecurity in plain English, and the lesson that teaches each one. - [Compare](https://insideout.sh/compare): Plain answers to the AI and quant engineering choices people search for, each leading to the lessons that teach it. - [Practice](https://insideout.sh/practice): Problems with fresh numbers every time: find your level, a short set each day, and three-minute speed runs. AI, quant and security. - [Interview](https://insideout.sh/interview): Practice for quant, AI engineering and security interviews, round by round: pick your role and your date, then practise each round. - [Projects](https://insideout.sh/projects): Hands-on projects for every module, built on your own machine with free data and your own tools. - [Blog](https://insideout.sh/blog): Long guides from insideout: how to prepare for quant and AI interviews, and how models and markets work, one step at a time. - [Pricing](https://insideout.sh/pricing): Pro is $9.99 a month, $23.97 every three months, $71.88 a year ($5.99 a month) or $119 once for life. These are US prices; other countries see a local price. ## AI engineering - [Classical Machine Learning](https://insideout.sh/courses/classical-ml): Fit models to tables of data, and measure them honestly. - [Neural Networks](https://insideout.sh/courses/neural-networks): Build a small network and train it with gradients you compute. - [Tokens and Embeddings](https://insideout.sh/courses/tokens-embeddings): How text becomes numbers a model can learn from. - [Transformers](https://insideout.sh/courses/transformers): Build the architecture behind every modern language model. - [Training and Tuning](https://insideout.sh/courses/training-tuning): How a base model is trained, then turned into an assistant. - [Inference and Serving](https://insideout.sh/courses/inference-serving): Make a model answer fast and cheaply under real traffic. - [Prompting](https://insideout.sh/courses/prompting): Get reliable, well-formed answers out of a model. - [Evaluation](https://insideout.sh/courses/evals): Measure whether an AI system is actually getting better. - [Search and Retrieval](https://insideout.sh/courses/search-retrieval): Find the right text and ground a model's answer in it. - [Agents](https://insideout.sh/courses/ai-agents): Let a model act with tools, inside limits you set. - [Production AI](https://insideout.sh/courses/production-ai): Run AI features people can depend on. - [Engineering Toolkit](https://insideout.sh/courses/engineering-toolkit): The working basics that AI and quant teams expect you to know. Optional. ## Quant engineering - [Probability](https://insideout.sh/courses/probability): The maths of chance that every quant interview starts with. - [Statistics and Regression](https://insideout.sh/courses/statistics): How to learn from noisy data without fooling yourself. - [Time Series](https://insideout.sh/courses/time-series): How market data behaves over time, and what it can predict. - [Python for Market Data](https://insideout.sh/courses/python-data): Working with market data in numpy and pandas, without look-ahead. - [Market Microstructure](https://insideout.sh/courses/markets): How an exchange matches orders, and how market makers set prices. - [Portfolio and Risk](https://insideout.sh/courses/portfolio): How to combine bets and measure the risk of the whole. - [Puzzles and Brainteasers](https://insideout.sh/courses/puzzles): The methods behind the logic puzzles in quant interviews. Optional. ## Cybersecurity - [Computing Foundations](https://insideout.sh/courses/computing-foundations): How computers, operating systems and code work, as security sees them. - [Networking Foundations](https://insideout.sh/courses/networking-foundations): How data crosses a network, layer by layer, and what each layer trusts. - [Security Foundations](https://insideout.sh/courses/security-foundations): How real breaches start, and the principles that would stop them. - [Cryptography](https://insideout.sh/courses/cryptography): What each primitive guarantees, and how to run crypto in production. - [Identity and Access](https://insideout.sh/courses/identity-access): Who may do what, checked on every request, for people and machines. - [Network Security](https://insideout.sh/courses/network-security): Networks designed so one foothold can't reach everything. - [Systems and Endpoint Security](https://insideout.sh/courses/systems-endpoints): Hosts hardened, patched and watched, down to the firmware. - [Cloud Security](https://insideout.sh/courses/cloud-security): Cloud accounts and workloads set up so one mistake stays small. - [Security Operations](https://insideout.sh/courses/security-operations): Watch the alerts, judge what's real, and escalate the right ones. - [Detection and Response](https://insideout.sh/courses/detection-response): Write detections that hold, and run the incident that follows. - [Privacy and Data Protection](https://insideout.sh/courses/privacy-data): Know what personal data you hold, keep less of it, protect the rest. - [Governance, Risk and Compliance](https://insideout.sh/courses/governance-risk): Run security as decisions with owners, evidence and deadlines. - [Web and API Security](https://insideout.sh/courses/web-api-security): Web and API flaws by root cause, with the fix that removes each. - [Secure Development](https://insideout.sh/courses/secure-development): Find security flaws before release, and fix the worst ones first. - [Software Supply Chain](https://insideout.sh/courses/supply-chain): Prove where every package and build came from before it runs. - [Security Assessment](https://insideout.sh/courses/security-assessment): Test systems the way an attacker would, with permission and a report. - [Threat Intelligence](https://insideout.sh/courses/threat-intelligence): Turn what's known about attackers into decisions your team can use. Optional. - [Digital Forensics](https://insideout.sh/courses/digital-forensics): What evidence can prove, how to keep it sound, and how to report it. Optional. - [OT and ICS Security](https://insideout.sh/courses/ot-ics-security): Secure plants and devices where safety and uptime come first. Optional. - [Risk, Audit and Leadership](https://insideout.sh/courses/risk-audit-leadership): Run risk, audits and a security function the way leaders are judged. Optional. ## Comparisons - [RAG vs fine-tuning](https://insideout.sh/compare/rag-vs-fine-tuning): When to give a model the right documents at question time, when to change its weights, and when a product needs both. - [Prompting vs fine-tuning](https://insideout.sh/compare/prompting-vs-fine-tuning): When a better prompt will do, when to train the model on your own examples, and how to tell which problem you have. - [Vector search vs keyword search](https://insideout.sh/compare/vector-search-vs-keyword-search): Embeddings against BM25: where each one finds the right passage, where each one misses, and how hybrid search merges the two lists. - [Agents vs workflows](https://insideout.sh/compare/agents-vs-workflows): When to let a model choose each next step, when to fix the path in code, and why most teams should start with the simpler one. - [MCP vs function calling](https://insideout.sh/compare/mcp-vs-function-calling): What function calling and the Model Context Protocol each do, how they meet in one tool call, and when an MCP server is worth building. - [LLM-as-judge vs human graders](https://insideout.sh/compare/llm-as-judge-vs-human-graders): When a model can grade your AI system's answers, when you need people, and how to check the judge against human marks. - [Quantization vs a smaller model](https://insideout.sh/compare/quantization-vs-smaller-model): When to store a model's weights in 8 or 4 bits, when to switch to a smaller model, and how to check what each one costs you in quality. - [Mixture of experts vs dense models](https://insideout.sh/compare/mixture-of-experts-vs-dense): How a mixture-of-experts model differs from a dense one in memory, compute per token and serving cost, and how to read names like 235B-A22B. - [AI engineer vs ML engineer](https://insideout.sh/compare/ai-engineer-vs-ml-engineer): What AI engineers and ML engineers do day to day, the skills each role needs, how their interviews differ, and where the two overlap. - [Simple vs log returns](https://insideout.sh/compare/simple-vs-log-returns): The two ways to measure a return, when each one is right, how to convert between them, and why the gap matters for volatile assets. - [Quant researcher vs quant developer](https://insideout.sh/compare/quant-researcher-vs-quant-developer): What quant researchers and quant developers do day to day, the skills and interviews each needs, and how to choose. - [Quant trader vs quant researcher](https://insideout.sh/compare/quant-trader-vs-quant-researcher): What quant traders and quant researchers do day to day, the maths each needs, how the interviews differ, and how to choose. ## Interview prep - [Quant Trader interview prep](https://insideout.sh/interview/quant-trader): What each round of a Quant Trader interview is like, with practice for each round and a plan to your date. - [Quant Researcher interview prep](https://insideout.sh/interview/quant-researcher): What each round of a Quant Researcher interview is like, with practice for each round and a plan to your date. - [Quant Developer interview prep](https://insideout.sh/interview/quant-developer): What each round of a Quant Developer interview is like, with practice for each round and a plan to your date. - [AI Engineer interview prep](https://insideout.sh/interview/ai-engineer): What each round of an AI Engineer interview is like, with practice for each round and a plan to your date. - [ML Engineer interview prep](https://insideout.sh/interview/ml-engineer): What each round of an ML Engineer interview is like, with practice for each round and a plan to your date. - [Inference Engineer interview prep](https://insideout.sh/interview/inference-engineer): What each round of an Inference Engineer interview is like, with practice for each round and a plan to your date. - [SOC Analyst](https://insideout.sh/interview/soc-analyst): For analysts who triage alerts, read logs and escalate the real ones. The lessons that prepare you for it, in order. - [GRC and Risk Analyst](https://insideout.sh/interview/grc-risk-analyst): For analysts who keep risk, controls and audits in order. The lessons that prepare you for it, in order. - [Mental maths test](https://insideout.sh/interview/mental-math): A free mental maths test in the format of trading firms' numerical tests: an 8-minute clock, multiple choice, and a point off for each wrong answer. - [Probability problems](https://insideout.sh/interview/probability): Original probability and expected value problems for quant interviews, with fresh numbers, hints and worked solutions. The first level of each is free. ## Blog - [How large language models work, step by step](https://insideout.sh/blog/how-llms-work): How a large language model turns text into tokens, tokens into vectors and vectors into the next word, with every step worked by hand. Markdown: https://insideout.sh/blog/how-llms-work.md - [How to prepare for a quant interview](https://insideout.sh/blog/quant-interview-prep): What quant trader, researcher and developer interviews test, round by round, with worked examples and an eight-week plan you can follow. Markdown: https://insideout.sh/blog/quant-interview-prep.md - [Probability questions in quant interviews, worked step by step](https://insideout.sh/blog/quant-interview-probability-questions): Ten probability problems of the kind quant trading and research interviews use, each worked step by step, with the method that solves it. Markdown: https://insideout.sh/blog/quant-interview-probability-questions.md - [Mental maths for trading interviews: methods that make you fast](https://insideout.sh/blog/mental-math-trading-interviews): The mental arithmetic that trading interviews and numerical tests expect, the shortcuts that make it fast, and a ten-minute daily routine. Markdown: https://insideout.sh/blog/mental-math-trading-interviews.md - [How to become an AI engineer: what to learn, in order](https://insideout.sh/blog/ai-engineer-roadmap): What an AI engineer does, what to learn and in what order, how long each part takes, and how to show you can do the job. Markdown: https://insideout.sh/blog/ai-engineer-roadmap.md ## Optional - [All public text in one file](https://insideout.sh/llms-full.txt): every module's outline, every free lesson in full, the comparisons and the glossary. - [New lessons feed](https://insideout.sh/feed.xml): RSS. - [Sitemap](https://insideout.sh/sitemap.xml): every page search engines may list, with its last update. - Markdown copies: each module page and each free lesson has one at the same address plus .md, for example https://insideout.sh/courses/classical-ml/01.md