--- title: "Managing documentation projects with AI course" id: "18238" type: "page" slug: "managing-documentation-projects-ai" published_at: "2025-09-25T10:55:28+00:00" modified_at: "2026-10-05T06:14:30+00:00" url: "https://www.cherryleaf.com/training-courses/managing-documentation-projects-ai/" markdown_url: "https://www.cherryleaf.com/training-courses/managing-documentation-projects-ai.md" excerpt: "Build, test, and decide: put AI to work in your documentation process A self-paced, assessed programme for documentation managers and senior writers. Build no-code AI workflows, test them against real cases, and make an evidence-based recommendation your organisation can act..." --- ## Build, test, and decide: put AI to work in your documentation process **A self-paced, assessed programme for documentation managers and senior writers. Build no-code AI workflows, test them against real cases, and make an evidence-based recommendation your organisation can act on. Tutor feedback and certificate included.** - Self-paced online programme - Equivalent to a two and a half days classroom course - 13 months’ access - Individual assessment, written feedback and certificate - Early bird price: £695 per delegate ex VAT Along the way, you will build and test no-code applications and agents for tasks across the documentation lifecycle. You will finish with an evidence-based recommendation to pilot, implement, revise, restrict, or stop an AI-assisted documentation workflow. [Buy this course](https://checkout.teachable.com/secure/55566/checkout/order_7ztn0l0g) ## What you will produce During the programme, you will select one documentation workflow and examine whether AI or conventional automation could improve it. You will produce: - A clearly defined opportunity and pilot boundary - A project plan - A workflow showing the fixed rules, AI-assisted decisions, and human approval points - An AI evaluation - A recommendation explaining whether the workflow should proceed You can base the project on your organisation’s work, use anonymised materials, or complete a fictional scenario supplied with the programme. ## What you will build Through the demonstrations and exercises, you can build no-code applications and agents covering the documentation lifecycle, working through them at your own pace. Each one shows where AI adds value, where fixed rules work better, and where a person needs to stay in control. ### Project management and reporting - Estimating app: produce effort estimates from project information - Staff scheduling app: allocate work across the team - Review assignment agent: match content to suitable reviewers - ROI dashboard: show the value of your documentation ### Researching and gathering information - Information gathering app: collect and organise source material - “What’s changed?” agent: detect product changes that affect documentation - Feature briefing pack: turn scattered evidence into a sourced briefing - Feedback analysis agent: find patterns in user feedback ### Information design and drafting - Metadata analyser and Metadata and tagging agent: assess and apply metadata - Diátaxis Auditor: check content against the Diátaxis framework - Content reuse agent: identify content that can be shared or reused - Audience persona adaptation agent: adapt approved content for different readers - Screen data anonymiser: remove personal data from screenshots ### Checking and publishing - Reader simulator: test how readers might respond to your content - Code example validator: check that code examples work - Security and privacy scan agent: flag sensitive information before publication - Pre-publication checklist: confirm required checks and approvals ### Maintaining content - Content lifecycle manager: track review dates and decide what to update or retire ## What you will learn AI can support documentation work at many points, from analysing source material to checking published content. The management task is deciding where it adds value, what evidence it can use, and where human responsibility must remain. By the end of the programme, you will be able to: - Build and test no-code AI applications and agents for tasks across the documentation lifecycle. - Identify a documentation problem that is suitable for AI or conventional automation. - Compare chatbots, AI assistants, no-code applications, coding agents, and automated workflows. - Define a focused pilot with clear boundaries, exclusions, responsibilities, and measures of success. - Decide which parts of a workflow need fixed rules, AI-assisted judgement, or a human decision. - Assess whether the available source information is reliable and sufficient for the work. - Design a controlled workflow with evidence, validation, approval points, permissions, monitoring, and safe failure behaviour. - Test AI performance using representative tasks, difficult cases, and defined acceptance criteria. - Address privacy, confidentiality, copyright, supplier, security, and governance requirements. - Estimate implementation effort and total cost of ownership. - Use evidence from the pilot to recommend whether the workflow should proceed, be revised, have its use restricted, or be stopped. - Plan how an approved workflow will be operated, monitored, maintained, and re-evaluated. The programme applies these decisions across the documentation lifecycle, including planning, research, information design, drafting, checking, publishing, maintenance, and measurement. ## Who this programme is for This programme is for people who need to make decisions about AI-assisted documentation work, including: - Documentation managers and heads of content - Lead Technical Authors and Technical Writers - Content operations and knowledge managers - Policy and procedure managers - Project managers responsible for documentation - Compliance and governance specialists involved in documentation workflows You do not need to be a programmer. You should have experience of documentation, policies, procedures, knowledge management, or content projects. ### How this differs from our other AI course We also offer a [Using Generative AI in Technical Writing](https://www.cherryleaf.com/training-courses/ai-technical-writing-course/) . e-learning course. The table below describes the difference between the courses: | Course | Best suited to | | --- | --- | | Using Generative AI in Technical Writing | People who want to use AI when researching, drafting, reviewing, publishing, and maintaining individual pieces of content | | Managing Documentation Projects with AI | People responsible for selecting, evaluating, governing, and implementing AI-assisted documentation workflows | Both courses share the same module on managing privacy, security, copyright, and governance. ## The final assessed project The final project gives you a structured way to apply the course to a documentation decision. You will choose one documentation activity and investigate whether AI, conventional automation, or a process change could improve it. You will design a controlled workflow, test it with representative examples, and make an evidence-based recommendation. Your project does not need to result in a recommendation to use AI. Identifying that a proposed workflow is too risky, too expensive, or unlikely to solve the problem is a valid outcome. ### Choose a project route You can complete the project in one of three ways: - Use a documentation process from your current organisation. - Use anonymised or fictionalised versions of workplace information. - Use the fictional organisation and source materials supplied with the course. The fictional route is suitable if you are not currently working for an organisation or cannot use workplace material for confidentiality reasons. All three routes are assessed against the same criteria. ### What you will do You will: - Define the documentation problem, intended benefit, pilot boundary, and exclusions. - Map the trigger, inputs, context, processing, controls, output, and monitoring. - Decide which stages need fixed rules, AI-assisted judgement, or a human decision. - Identify the approved sources, evidence requirements, permissions, and review responsibilities. - Test or simulate the workflow using routine, uncertain, high-risk, and failed cases. - Evaluate the results against measures such as accuracy, review effort, time saved, unsupported claims, and total cost. - Recommend whether the organisation should proceed, revise the design, restrict its use, or stop the project. You can use an AI assistant, coding agent, low-code platform, workflow tool, or a documented manual simulation. We assess the quality of your decisions and evidence, rather than the technical complexity of the prototype. ### What you will submit You will email us a report together with any useful supporting material. The supporting material might include: - A workflow diagram - A project plan - Sample inputs and outputs - Prompts or AI rules - A test record - A risk assessment - A roles and responsibilities document - A total cost of ownership assessment Do not send passwords, personal data, confidential business information, unpublished product information, or material you are not authorised to share. ### Feedback and certificate We will review your submission and send you written feedback. If your project meets the assessment standard, you will receive a certificate of successful completion. If your first submission needs further work, you will have one opportunity to revise and resubmit it. ## How the programme works The programme combines on-demand learning, practical exercises, an assessed project, and access to tutor support. ### Learn at your own pace The course modules are delivered online as a series of short video presentations. You can pause the recordings, repeat a demonstration, and return to the programme when it suits you. The taught content is equivalent to approximately two and a half days of training. You should allow additional time for the exercises and final project. You will have access to the programme for 13 months from the date you enrol. ### Build and test AI tools and workflows The recordings explain the concepts, decisions, and risks involved in managing AI-assisted documentation work. Separate demonstrations then show you how to build and use applications and agents with AI assistants, workflow tools, and conventional automation. Practical exercises help you apply the ideas to your own documentation process or to the fictional materials supplied with the course. You do not need to be a programmer. The demonstrations and exercises are where you build and experiment. The final project is where you show sound judgement: you choose one documentation activity, test a workflow with representative cases, and decide whether your organisation should go ahead. The exercises are for your own learning and are not assessed. ### Build your project as you progress You do not need to wait until the end of the programme to begin the final project. Project recordings guide you through each stage, from selecting an opportunity and defining the scope to testing the workflow and preparing your recommendation. You will receive supporting documents that you can complete as you progress. This reduces the amount of work needed at the end. ### Ask questions during the programme Monthly online office hours give you an opportunity to ask questions about the course, exercises, or project. You can attend the sessions that are useful to you. The office hours help you think through a decision or problem; they are not additional compulsory lessons. ### Submit your final project When you are ready, email your final project to Cherryleaf. We will assess it against the published project criteria and send you written feedback. If your work meets the required standard, you will receive a certificate of successful completion. ## What is included The course includes: - On-demand course recordings - Demonstrations and practical exercises - Thirteen months’ access - A final-project guide - Fictional organisation and source materials - A project plan - An AI evaluation - Eleven portable technical-writing `SKILL.md` files - Monthly online office hours - Individual assessment of your final project - Written tutor feedback - An opportunity to revise and resubmit - A certificate of successful completion The eleven portable `SKILL.md` files are reusable instructions that teach an AI assistant how to carry out documentation tasks such as evaluating documentation quality and suggesting practical fixes, writing API documentation, and drafting release notes. Email us to request a free preview of one module. ## Topics covered ### Choosing and defining an opportunity Start with a documentation problem rather than an AI product. You will learn how to: - Compare opportunities based on value, feasibility, available evidence, ability to evaluate the result, and risk. - Define a focused pilot boundary. - State what the proposed workflow will not do. - Decide whether AI, conventional automation, or a process change is the better response. - Use Lean techniques to identify delays, rework, duplication, and other wasted effort. ### Understanding automation, AI assistants, and agents The programme distinguishes between predictable automation, generative AI, and agentic workflows. You will consider where fixed rules work best, where AI interpretation adds value, and where a person must remain responsible. You will compare: - Chatbots, AI assistants, and agent modes - No-code AI applications - Locally installed assistants and coding agents - Automated workflows that connect AI with other systems The programme examines their capabilities, costs, permissions, maintenance requirements, and failure modes. It also introduces tools, MCP connections, reusable skills, agent harnesses, and the chaining of smaller tasks. ### Planning the project and assigning responsibility Before automating a process, you need to understand how the work is performed today. You will learn how to: - Document the current process, including decisions, delays, exceptions, and handovers. - Define the intended outcome, scope, baseline, measures, resources, and timescale. - Identify who supplies information, reviews outputs, approves decisions, and owns the operational workflow. - Use AI to support estimating, dependency analysis, risk identification, staff scheduling, meeting analysis, and stakeholder reporting. ### Using AI across the documentation lifecycle The programme examines where AI and conventional automation can support each stage of a documentation project. #### Researching users, products, and source information - Summarise and compare source material. - Extract facts and identify missing information. - Analyse user feedback and support enquiries. - Track changes across source documents. - Decide whether the evidence is sufficient for writing to begin. - Record uncertainty rather than conceal it. #### Planning the scope - Identify content that should be created, updated, reused, or removed. - Compare the proposed scope with user needs and available evidence. - Identify dependencies, exclusions, and gaps. #### Information design - Explore content models, taxonomies, metadata, the Diátaxis framework, and content reuse. - Consider how structure affects findability, translation, accessibility, and AI retrieval. - Decide how content should be divided, labelled, linked, and maintained. #### Drafting content - Draft release notes, procedures, summaries, metadata, and alternative text. - Adapt approved information for different audiences. - Use screenshots and extracted interface information safely. - Preserve links between statements and their sources. #### Project management and reporting - Support estimating and work allocation. - Identify dependencies and project risks. - Prepare stakeholder updates. - Analyse meeting notes and project records. - Monitor progress without hiding uncertainty. #### Post-writing checks - Combine AI review with deterministic checks. - Check terminology, structure, links, code examples, accessibility, privacy, and compliance. - Prioritise findings instead of rewriting everything. - Define the evidence and approval needed before publication. #### Publishing - Prepare documentation for human readers and AI retrieval systems. - Improve the structure and metadata used by search, RAG systems, and AI assistants. - Retain suitable review and approval controls. #### Measuring whether documentation works - Define success and establish a baseline. - Use evidence from analytics, searches, support enquiries, feedback, and task completion. - Use AI to help identify patterns while leaving people responsible for interpreting the evidence. - Distinguish useful outcomes from measures of content production. ### Designing and automating a controlled workflow A documentation workflow contains more than a prompt. You will learn how to map: - The event that triggers the work - The permitted inputs - The context and rules - The transformation - The controls and validation - The output - The monitoring and notifications Worked patterns include: - Detecting product changes and proposing documentation updates - Generating draft release notes - Comparing planned functionality with what was built - Reviewing content against house rules - Testing documented procedures - Chaining small, reusable workflow steps You will consider scripts, repository workflows, low-code platforms, coding agents, scheduled AI tasks, and specialist documentation services. The programme also covers approval points, exception handling, sampling, thresholds, evidence links, audit records, and safe failure behaviour. ### Managing privacy, security, copyright, and governance An AI workflow can act on source material, connect to other systems, and produce convincing but unsupported content. Its permissions and controls matter. The programme covers: - Personal data and Data Protection Impact Assessments - Supplier contracts and records of processing - Confidentiality - Copyright, licensing, and intellectual property - AI decisions, transparency, bias, and user autonomy - Access controls and minimum permissions - API credentials and connected systems - Prompt injection and untrusted sources - Logging and incident preparation - Human oversight and accountability You will also consider how organisations can create practical AI rules that people can understand and apply. ### Governance and lifecycle management A workflow still needs an owner after the pilot ends. You will learn how to: - Assign operational and maintenance responsibility. - Record changes to instructions, sources, models, tools, and permissions. - Set review dates and audit requirements. - Detect changes that might affect the documentation. - Monitor output quality and workflow failures. - Prepare fallback and rollback arrangements. - Re-evaluate the workflow when an important component changes. ### Testing the workflow and preparing a recommendation You will test or simulate the workflow using representative cases. These should include: - Routine work - Incomplete or conflicting evidence - High-risk or excluded material - An untrusted source containing misleading instructions - A failed or incomplete workflow run You will define acceptance criteria before testing and record: - Unsupported claims - Missing information - Incorrect decisions - Unnecessary escalations - Human review effort - Time and cost - Failures correctly detected - Problems that escaped the controls You will use the evidence to recommend whether the organisation should proceed, revise the design, restrict its use, or stop. ### Workflow examples, tools, and demonstrations The programme includes documentation-workflow examples and case studies from organisations and services including Mintlify, Anthropic, Elastic, Payabli, Stripe, GitLab, Coca-Cola Andina, Skyflow, Cloudflare, and Microsoft Aspire. Demonstrations also examine approaches involving Claude Cowork, Claude Code, Claude Projects, OpenAI Codex, Microsoft Copilot, reusable skills, and goal-based agent workflows. Products will change. The programme concentrates on the decisions, evidence, and controls that remain necessary when the tools change. ## Our experience with AI-assisted documentation We began examining generative AI after speaking with developers who were researching its potential for technical writing. We looked at each stage of a documentation project, from planning and research to checking, publishing, and maintenance. We built and tested applications, agents, and automated workflows to find where AI was useful, where conventional automation worked better, and where human judgement remained essential. That work led to our AI training courses, conference presentations, and consultancy projects. This programme brings those lessons together into a method for planning, testing, and managing AI-assisted documentation work. Managing and mastering documentation projects with AI course ## Previous delegates and testimonials This self-paced programme builds on our instructor-led course. Previous live-course delegates have included staff from: - Aristocrat Interactive - Anthology - Bazaarvoice - Camunda - CData - Dell - Emplifi - Entrust - Forterro - ING - Instacart - Ivalua - Lendscape - Motorola Solutions - Opus 2 - Roche Diagnostics - SITA - Segpay Here is what delegates told us. ### Jill Lauterborn – Managing Documentation Projects with AI course Excellent class, in structure, delivery, and content! ### Anna Litvinskaya – Managing Documentation Projects with AI course I’ve attended the first edition of Managing and Mastering Documentation Projects with AI, and it was actually a great course. I highly recommend it to all technical writers with at least some experience writing documentation! (Posted on LinkedIn 17/1/25) Anna Litvinskaya CData Software ### AS – Managing Documentation Projects with AI course Great course, at just the right time for me. I’ve used AI to make the creation of individual docs more efficient but I was struggling to think about how tools could improve the management of a set of documentation. ### ZN – Managing documentation projects with AI The course was a perfect blend of discussion and practice. It increased my confidence using the AI tools and my retention of the content. ### Jean R – Managing documentation projects with AI I learned a lot and now have some practical experience that I can talk about and employ in my work. ### Evgenia B – Managing documentation projects with AI For me personally, it helped me break the psychological barrier to starting tackling AI and gave me the general understanding of how we can use AI in our processes. ### JB – Managing Documentation Projects with AI course This course did a great job of helping me uncover my own process blindness. As an experienced technical writer, it’s easy for me to forget the array of tasks & subtasks that go into the work I do because it’s practically second nature. So, it was refreshing to be able to sit still for a few hours, take notice of what I’ve tended to write off as “just part of the job,” and decide how best to handle the parts that still serve the project. ### AW – Managing Documentation Projects with AI course Overall, the course was well structured and put together. I found the practical exercises really helpful. ## Price and booking ### Early bird pricing Our introductory early bird launch price is £695 per delegate ex VAT. The post launch price is likely to be £895 per delegate, ex VAT. The price includes: - The complete on-demand programme - Thirteen months’ access - Demonstrations and practical exercises - Project planning and evaluation documents - A fictional project option - Monthly online office hours - Individual assessment of your final project - Written tutor feedback - An opportunity to revise and resubmit - A certificate of successful completion The programme is designed to help organisations avoid two expensive mistakes: investing in an AI workflow that does not solve a useful problem, and introducing automation without sufficient evidence, controls, or operational ownership. [Buy this course](https://checkout.teachable.com/secure/55566/checkout/order_7ztn0l0g) ## Private training and implementation support ### Private instructor-led courses We can deliver the course privately to a group within your organisation using Microsoft Teams or another agreed online meeting platform. Private delivery includes trainer-led explanations, demonstrations, exercises, and group discussions. Participants can compare their approaches and discuss how the principles apply to their organisation. We can adapt the emphasis to your documentation environment, industry, software platforms, and governance requirements. Subject to suitable security and confidentiality arrangements, exercises can use examples based on your organisation’s documentation processes. [Contact us](https://www.cherryleaf.com/contact_us/) to discuss group size, dates, tailoring, and price. ### Implementation support For organisations ready to put an AI-assisted documentation workflow into operation, we can provide support after the training. This can include: - 30-, 60-, or 90-day support through Slack or Microsoft Teams - Follow-up coaching calls - Help refining project plans, controls, and measures - Access to agent and workflow templates ### Done-with-you consultancy We can also work with your team to design and implement documentation workflows. This can include: - A discovery workshop - Prioritisation of suitable opportunities - Workflow and agent design - Development of agreed applications or agents - Training for the people who will operate and maintain them - Ongoing implementation support ## Frequently asked questions ### How much time should I allow? The taught content is equivalent to approximately two and a half days of training. You should allow additional time to assess and measure your current processes. You can divide the work into shorter sessions and complete it at your own pace. ### How long will I have access? You have access to the programme for 13 months from the date you enrol. ### Do I need programming experience? No. You can complete the exercises and project using an AI assistant, a no-code platform, a workflow tool, or a documented manual simulation. The course also introduces coding agents and repository-based workflows for learners who want to explore them. ### What if I cannot use information from my employer? You can complete the course using supplied fictional data and scenarios. It includes representative source materials, uncertainty, exclusions, and workflow risks. ### Can I take the programme if I am not currently employed? Yes. Use the fictional project route and treat the supplied organisation as a client or employer. ### Do I have to recommend using AI? No. A recommendation to use conventional automation, revise the proposal, restrict its use, or stop the project can be a successful outcome. ### What certificate will I receive? You will receive a certificate of successful completion after your final project meets the assessment standard. Watching the recordings without completing the project does not qualify for the assessed certificate. ### Are the monthly office hours compulsory? No. They give you an opportunity to ask questions about the course, exercises, or project. Attend the sessions that are useful to you. ### Can an organisation purchase several places? Yes. Contact us to discuss multiple licences, invoicing, progress reporting, or private delivery. ## Contact us Email [info@cherryleaf.com](mailto:info@cherryleaf.com) if you have a question about the programme, private training, or implementation support. We can also arrange a conversation using Microsoft Teams, Zoom, or another agreed online meeting platform.