# Linear Horizon

> Linear Horizon helps organisations understand, design and deploy artificial intelligence — through training, consultancy, production engineering and our own agentic products.

We teach AI, design it, build it, and provide the technology to run it. Contact hello@linearhorizon.com. Principal: Nicholas Johnson, AI architect and software engineer. HTML pages are canonical. Structured JSON for ingest is at https://linearhorizon.com/api/v1.json.

## Company

- [Home](https://linearhorizon.com/): From complexity to outcome.
- [About](https://linearhorizon.com/about): Who Linear Horizon is, and how training, consulting, build and products fit together.
- [Training](https://linearhorizon.com/training): Practitioner-led course catalogue for engineering teams, technical leaders and organisations.
- [Consulting](https://linearhorizon.com/consulting): Where AI is genuinely useful, and how to design it so it can be built.
- [Build](https://linearhorizon.com/build): Production agents, RAG and bespoke systems — or a Linear Horizon product deployment.
- [Products](https://linearhorizon.com/products): Platform, then specialist applications for analysis, research, strategy, growth and service.
- [Work](https://linearhorizon.com/work): Selected client projects, described by problem and outcome.
- [Insights](https://linearhorizon.com/insights): Notes from the practice on applied AI and putting systems into production.
- [Whitepapers](https://linearhorizon.com/whitepapers): Longer pieces on agents, adoption and organisational systems.
- [Business cases](https://linearhorizon.com/cases): Problem briefs: the situation, what fails, and a path through products and services.
- [Compare](https://linearhorizon.com/compare): How Claude Cowork, Grok Bot and Linear Horizon differ: a general agent is not an organisational system.
- [Contact](https://linearhorizon.com/contact): Talk to Linear Horizon, or email hello@linearhorizon.com.

## Products

- [Linear Horizon Platform](https://linearhorizon.com/products/platform): An operating system for AI.
- [Linear Horizon Stack](https://linearhorizon.com/products/stack): Ask open-ended questions of your data.
- [Linear Horizon Solve](https://linearhorizon.com/products/solve): Start a run. The tree is the ledger.
- [Linear Horizon Wargame](https://linearhorizon.com/products/wargame): A machine that can predict the future.
- [Linear Horizon Funnel](https://linearhorizon.com/products/funnel): Your data never leaves your desktop.
- [Linear Horizon Callcenter](https://linearhorizon.com/products/callcenter): One conversation, across every channel.
- [Linear Horizon Query](https://linearhorizon.com/products/query): Understand what users actually do.
- [Linear Horizon Chat](https://linearhorizon.com/products/chat): A conversation on top of your agents.

## Training programmes

- [AI Engineering with Python](https://linearhorizon.com/training/ai-engineering): Working agents, retrieval and production systems in Python.
- [Applied NLP with Python](https://linearhorizon.com/training/applied-nlp): Classical NLP through transformers, in Python.
- [AI Engineering with Python lab handbook](https://linearhorizon.com/training/ai-engineering/labs): https://github.com/nicholas-johnson/ai-with-python-course
- [Applied NLP with Python lab handbook](https://linearhorizon.com/training/applied-nlp/labs): https://github.com/nicholas-johnson/nlp-detective-agency

## Training topics

- [Topics](https://linearhorizon.com/training/topics): Course catalogue grouped by subject.
- [Foundations](https://linearhorizon.com/training/topics/foundations): Build an accurate picture of how generative AI works, and how to use it with judgement.
- [Leadership](https://linearhorizon.com/training/topics/leadership): Architectural, commercial and organisational decisions about AI that you can stand behind.
- [AI Engineering](https://linearhorizon.com/training/topics/ai-engineering): Build, test, secure and operate applications that use large language models.
- [RAG & Agents](https://linearhorizon.com/training/topics/rag-agents): Design retrieval systems, tool-using agents, and the combination of the two.
- [Architecture](https://linearhorizon.com/training/topics/architecture): Give AI a production shape: platforms, assistants, security boundaries and the operational work that follows.
- [Models & Machine Learning](https://linearhorizon.com/training/topics/models-ml): How models work, how they are adapted, and how they are run in practice.
- [Cloud & Platform](https://linearhorizon.com/training/topics/cloud-platform): Cloud implementations, and the internal platforms that make AI usable across your teams.
- [Workshops](https://linearhorizon.com/training/topics/workshops): Focused build days and facilitated working sessions around a problem your team actually has.
- [Audiences](https://linearhorizon.com/training/for): Catalogue filtered for engineers, architects and leaders.
- [AI Training for Software Engineers](https://linearhorizon.com/training/for/software-engineers): Practitioner-led AI training for software engineers: engineering fundamentals, RAG, agents, evaluation, security and production architecture.
- [AI Training for CTOs and Technical Leaders](https://linearhorizon.com/training/for/technical-leaders): AI training for CTOs, engineering managers and senior technical leaders: architecture choices, vendors, risk, and a first production path.
- [AI Training for Business Leaders and Directors](https://linearhorizon.com/training/for/business-leaders): A concise set of AI courses for senior leaders and directors: where the technology creates value, where it does not, and how to judge a first serious step.
- [AI Architecture Training for Enterprise Teams](https://linearhorizon.com/training/for/architects): AI architecture training for enterprise architects and senior engineers: platforms, assistants, retrieval, cloud implementations and operating constraints.

## Courses

- [Generative AI Fundamentals](https://linearhorizon.com/training/generative-ai-fundamentals): A technically accurate, practical introduction to modern generative AI for mixed technical and non-technical audiences. Enough depth to make sound decisions, with no requirement to write code.
- [Practical Generative AI at Work](https://linearhorizon.com/training/practical-generative-ai-at-work): A working day for knowledge workers and business teams: using generative AI for research, writing, analysis and document work, with enough discipline to trust the output.
- [How Large Language Models Work](https://linearhorizon.com/training/how-large-language-models-work): A one-day explanation of LLM internals for developers, architects and technically curious professionals. Enough mechanism to improve engineering decisions, without a research-paper prerequisite.
- [AI for Technical Leaders](https://linearhorizon.com/training/ai-for-technical-leaders): A decision-making course for CTOs, engineering managers, architects and senior technical product leaders. How to choose architectures, vendors and investments that will still make sense in production.
- [AI for Business Leaders](https://linearhorizon.com/training/ai-for-business-leaders): A concise course for senior leaders and directors: where AI creates actual value, where it does not, and how to judge investment, risk and organisational readiness.
- [AI-Native Programme Management with Claude](https://linearhorizon.com/training/ai-native-programme-management-with-claude): A 90-minute seminar for programme and project leaders: how to move from individual Claude use to connected, governed workflows across the programme lifecycle.
- [AI Engineering for Software Developers](https://linearhorizon.com/training/ai-engineering-for-software-developers): A three-day programme for professional software engineers: how to build real applications with modern AI, from model APIs and RAG through agents, evaluation and production patterns.
- [Building Applications with Generative AI](https://linearhorizon.com/training/building-applications-with-generative-ai): A two-to-three-day build course for software engineers: take an AI-powered application from architecture through RAG, tools, evaluation and deployment, ending with an end-to-end system.
- [Evaluating and Testing LLM Applications](https://linearhorizon.com/training/evaluating-and-testing-llm-applications): A one-day course on evaluating non-deterministic systems: golden datasets, semantic checks, LLM-as-judge, retrieval and agent evaluation, and the pipeline that keeps quality from drifting.
- [Securing Generative AI Applications](https://linearhorizon.com/training/securing-generative-ai-applications): A one-day security course for developers, security engineers and architects: the specific risks created by LLM applications, and the controls that actually reduce them.
- [RAG: From Fundamentals to Production](https://linearhorizon.com/training/rag-from-fundamentals-to-production): A two-day course on retrieval-augmented generation as an engineering discipline: why naive RAG fails, and how to design a system that can be evaluated, secured and operated.
- [Building AI Agents](https://linearhorizon.com/training/building-ai-agents): A two-day course for intermediate and advanced developers: design and build useful agents with tools, state, permissions and evaluation, rather than unsupervised demos.
- [Agentic RAG Systems](https://linearhorizon.com/training/agentic-rag-systems): A two-day advanced course combining retrieval with agentic reasoning and tool use: multi-step retrieval, query planning, structured sources and a production architecture for complex knowledge work.
- [Production AI Architecture](https://linearhorizon.com/training/production-ai-architecture): A two-day architecture course for senior engineers and technical leaders: reference architectures, model gateways, RAG, agents, security, evaluation and the operational concerns that decide whether a system can leave the lab.
- [Building Enterprise AI Assistants](https://linearhorizon.com/training/building-enterprise-ai-assistants): A two-day course for developers and platform teams: build a secure internal assistant that can use organisational knowledge and systems, with identity, permissions, citations and evaluation.
- [Architecting Enterprise Generative AI](https://linearhorizon.com/training/architecting-enterprise-generative-ai): A two-day course for architects and senior technical leaders: design enterprise-scale AI capability as a platform, not as a sequence of isolated projects.
- [Transformers from First Principles](https://linearhorizon.com/training/transformers-from-first-principles): A two-day course for technically strong practitioners: build a usable understanding of transformer models by implementing simplified components, without unnecessary mathematical formalism.
- [Fine-Tuning Language Models](https://linearhorizon.com/training/fine-tuning-language-models): A two-day course for ML engineers and experienced AI developers: when to adapt a model, how to do it with parameter-efficient methods, and how to evaluate and deploy the result.
- [Training Your Own Language Model](https://linearhorizon.com/training/training-your-own-language-model): A two-to-three-day course for advanced developers and ML engineers: build and train a small language model in order to understand the process, not to compete with frontier systems.
- [Running LLMs Locally](https://linearhorizon.com/training/running-llms-locally): A one-day practical course on local and private inference: model formats, hardware, quantisation, engines, benchmarking and the privacy and cost reasons to run models yourself.
- [AI Systems on AWS](https://linearhorizon.com/training/ai-systems-on-aws): A two-day course for cloud engineers, developers and architects: design and build AI systems on AWS, with architecture first and managed services in their proper place.
- [AI Systems on Azure](https://linearhorizon.com/training/ai-systems-on-azure): A two-day course for cloud engineers, developers and architects: design enterprise AI systems on Azure, including identity, search, model APIs and the architecture those services should sit inside.
- [AI Platform Engineering](https://linearhorizon.com/training/ai-platform-engineering): A two-day course for platform engineers, architects and senior developers: build the internal infrastructure that lets multiple product teams use AI safely and consistently.
- [RAG in a Day](https://linearhorizon.com/training/rag-in-a-day): A focused workshop for developers: build a complete retrieval-augmented generation system in one day, from ingestion and chunking through citations and a first evaluation.
- [Build an AI Agent in a Day](https://linearhorizon.com/training/build-an-ai-agent-in-a-day): A hands-on workshop for developers: build a working agent with tool calling, state, structured outputs, error handling, permissions and a first evaluation.
- [From Chatbot to Agent](https://linearhorizon.com/training/from-chatbot-to-agent): A one-day workshop that starts with a basic LLM chat application and evolves it, step by step, into a useful agentic system with retrieval, tools, state and production safeguards.
- [AI Architecture Workshop](https://linearhorizon.com/training/ai-architecture-workshop): A facilitated workshop for technical teams, architects and engineering leaders around a real customer problem. The output is an architecture and a next-step roadmap, not a training certificate.
- [AI Use-Case Discovery Workshop](https://linearhorizon.com/training/ai-use-case-discovery-workshop): A discovery workshop for business and technical leaders: identify where AI can create meaningful value, score the opportunities, and leave with a ranked list and recommended next steps.

## Insights

- [What is an agent? What does it mean to be agentic?](https://linearhorizon.com/insights/what-is-an-agent): An agent is an LLM, plus a prompt, plus tools — placed inside a loop that lets it pursue a goal. To be agentic is to give a system limited discretion over how that goal is achieved.
- [The gap between learning about AI and putting it into production](https://linearhorizon.com/insights/from-learning-to-production): Training without a path into the system leaves organisations fluent and stuck. The work is to close that gap.
- [Where AI is actually useful — and where it is not](https://linearhorizon.com/insights/where-ai-is-useful): The useful question is not what a model can do in general. It is what it can do here, against this data, these interfaces, and this risk.

## Whitepapers

- [The product as its own expert](https://linearhorizon.com/whitepapers/the-product-as-its-own-expert): Give each product an MCP server so an agent can inspect its code, telemetry, usage and history — then diagnose problems, cite evidence and propose the next improvement.
- [How Silicon Valley is approaching fully agentic workflows](https://linearhorizon.com/whitepapers/how-silicon-valley-is-approaching-fully-agentic-workflows): Leading technology companies are not making entire organisations autonomous. They are making particular workflows agentic — with a defined objective, connected context, bounded authority and deliberate human control.
- [The four stages of AI adoption in the workplace](https://linearhorizon.com/whitepapers/four-stages-of-ai-adoption): A practical four-stage model for AI adoption — Assist, Standardise, Connect and Transform — and how to progress without treating autonomy as the goal.

## Work

- [Linear Horizon ITAD Assist](https://linearhorizon.com/work/itad-assist): Find the best next life for every retired IT asset.
- [RoboRigs](https://linearhorizon.com/work/roborigs): From a sentence to a 3D-printable robot.
- [SukiTutor](https://linearhorizon.com/work/tabbytutor): A homeschool tutor for curious minds.

## Business cases

- [A conversation on top of the system, not instead of it](https://linearhorizon.com/cases/interface-to-agents): Chat is how people talk to agents. It is not the agent system. The case is an interface that shows working, respects identity, and leaves the work to the runtime.
- [A general agent is not an organisational system](https://linearhorizon.com/cases/organisational-systems): Handing work to a general agent is not the same as running a system other people can operate. The difference is inspectable state, specialist products and human gates.
- [Ask the question of your data](https://linearhorizon.com/cases/investigation): Dashboards report. Investigation forms hypotheses, tests them against evidence, and produces an account you can argue with.
- [Capability inside the organisation](https://linearhorizon.com/cases/capability): Practitioner-led training is for teams who have to ship, not for a public timetable of AI awareness.
- [Find the companies that need you](https://linearhorizon.com/cases/who-to-contact): The expensive part of prospecting is not sending a message. It is knowing who is worth contacting, and why — without sending that research to a cloud CRM.
- [From complexity to a course of action](https://linearhorizon.com/cases/complexity): Scattered work, tools and decisions do not become a system by adding a model. The case is to turn complexity into a path an organisation can actually operate.
- [From experimentation to operating practice](https://linearhorizon.com/cases/operating-practice): Adoption is not a workshop. It is the move from trials that impress to a way of working the organisation can repeat.
- [One customer, one conversation](https://linearhorizon.com/cases/one-conversation): Service fails when email, voice and messaging are treated as separate jobs. The case is a single journey with context, permitted actions and a human handoff that does not start again.
- [Search a hard problem, do not declare it solved](https://linearhorizon.com/cases/hard-problems): Some claims need an unattended search across programmes of attack, with a kernel that will refuse an unearned proof.
- [See what users do, then investigate why](https://linearhorizon.com/cases/what-users-do): Analytics that only report leave you with a number and no method. Measure behaviour, then let an agent pursue the question a dashboard cannot.
- [Strategy under uncertainty](https://linearhorizon.com/cases/strategy): The future is a distribution of interacting actors, not a point forecast. Wargame makes the assumptions explicit and explores their consequences.
- [Taking AI into production](https://linearhorizon.com/cases/production): A demonstration is not a system. Production means owners, interfaces, failure modes and a path out of the lab.
- [When a product is not the whole answer](https://linearhorizon.com/cases/bespoke): Platform and the specialist applications cover a lot of ground. Some estates, identity models and failure modes still need a system designed around them.
- [Why an operating system, not a chatbot](https://linearhorizon.com/cases/operating-system): Chat is an interface. An operating system for AI is what lets agents use tools, knowledge and workflows — and stop when they should.
- [Work out what to build — before you build it](https://linearhorizon.com/cases/discovery): The expensive mistake is applying AI indiscriminately. Discovery is whether the problem is real, bounded, and worth a system.

## Optional

- [JSON catalog](https://linearhorizon.com/api/v1.json): Index of every public JSON file.
- [JSON bundle](https://linearhorizon.com/api/v1/all.json): Single-file dump of public site content.
- [Courses JSON](https://linearhorizon.com/api/v1/courses.json): Full course catalogue as JSON.
- [Products JSON](https://linearhorizon.com/api/v1/products.json): Product family as JSON.
- [Sitemap](https://linearhorizon.com/sitemap-index.xml): HTML sitemap for crawlers.
- [Horizon-Recruiter](https://recruiter.linearhorizon.com/): Desktop CRM for recruiters.
- [Horizon Training Provider](https://trainingprovider.linearhorizon.com/): Desktop CRM for training providers.
- [Funnel app](https://app.linearhorizon.com/): Accounts, trial, and download.