AI Chat Agents
Support, sales and internal assistants, built and run for you.
Alongside the phases · Any stage
AI Training
Practical AI training, taught by people who build AI systems every day. We write and record video courses for institutes and course providers, run live workshops, and train teams on their own tools. Learners watch a working build from start to finish, errors included, then build their own.
- Modules taught
- 8
- Formats
- 3
- Needed first
- No audit
- Custom AI assistants
- AI in your workspace
- AI images and documents
- AI video and avatars
- APIs and webhooks
- No-code automations
- AI agents
- Vibe coding
What does AI Training include?
AI Training from Hyperloop AI includes lessons or sessions built to your syllabus or your team's goals, written guidelines for each lesson, supporting resources such as sample knowledge bases, guides and spreadsheets, demo businesses so no real data appears, reasonable revisions and a handover in the format your platform needs. Who owns the content is agreed up front.
AI Training at a glance
- Formats
- Recorded video courses · Live workshops · Team training on your own tools and use cases
- Who it's for
- Institutes, academies and course providers that need AI curriculum content, and companies upskilling their teams
- Topics
- Custom AI assistants · AI in your workspace and AI browsers · AI images, design and documents · AI video and avatars · APIs and webhooks · No-code automations · AI agents · Vibe coding
- Style
- Screen-recorded, workshop-style builds, with only as much theory as each build needs
- Comes with
- Written guidelines per lesson, plus sample knowledge bases, guides and spreadsheets
- Built to
- Your syllabus or programme outline, including module topics and minimum video lengths
- Revisions
- Reasonable revisions on request
- Ownership
- Agreed up front. Recorded courses can transfer in full to you as work for hire
Who is it for?
AI Training is for three groups: institutes and course providers that need practical AI content for a programme or certification, companies upskilling a team that has AI tools but rarely uses them beyond a chat window, and clients we're already building for whose team needs to run, adapt and extend the systems themselves.
Institutes and course providers
You set the outline. We write the lessons, build the demos, record them and hand over the videos with written guidelines and resources, ready to publish on your learning platform under your name.Companies upskilling their teams
Your team has AI subscriptions but still copies and pastes between chat windows. We train them on the tools you already pay for, using your own processes as the examples.Clients we're building for
Your team learns to run, adjust and extend what we've built, and to build the smaller things themselves.
Which format fits?
AI Training comes in three formats. A recorded video course suits self-paced programmes, certifications and training you want to reuse with every cohort. A live workshop suits cohorts, bootcamps and team away days. Team training suits companies rolling out AI across a team, using your own tools and documents as the examples.
Illustrative
- Lesson 1Recorded
- Lesson 2Recorded
- GuidelinesIncluded
01 of 03
Screen-recorded lessons you can reuse
Screen-recorded lessons, each with written guidelines, plus shared resources such as a sample company knowledge base. Good for self-paced online programmes and certifications, and for training you want to reuse with every new cohort or starter.
Would you rather we built it for you?
Teams that want a system working without building it themselves can move straight to AI Development. Many businesses do both.
AI Voice Agents
Phone agents that answer and make calls, built and run for you.
AI Automations
Workflows that capture, qualify and follow up, built and run for you.
How do we teach?
Hyperloop AI teaches by building. Lessons show a working build live, mistakes and fixes included, using made-up demo businesses rather than real client data. Materials carry from one lesson to the next, the method is taught alongside the buttons, technical ideas get everyday analogies, and lessons say plainly where a shortcut stops being production-ready.
- Error 400Fixing
- RetrySuccess
01 of 05
Build live, keep the mistakes in
Errors, outages and wrong settings are fixed on camera. Learners see how to read an error and recover, which is what they'll face on their own builds.
What can we teach?
Hyperloop AI teaches eight modules, each written, built and recorded as part of a delivered course: custom AI assistants, AI in your workspace and AI browsers, AI images and documents, AI video and avatars, APIs and webhooks, no-code automations, AI agents and vibe coding. Examples, depth and tools are tailored to your learners.
| Module | What learners can do afterwards |
|---|---|
| Custom AI assistants | Turn a general chat model into an assistant for one job, with a structured prompt and a well-organised knowledge base, in Gemini Gems, ChatGPT Custom GPTs and Projects, or Claude Projects |
| AI in your workspace and AI browsers | Use AI inside Google Docs, Sheets, Gmail and Drive, and use an AI browser for sourced research and agentic tasks |
| AI images, design and documents | Write structured image prompts, generate and refine images, keep a prompt library, and turn a brief into a deck or social post with Gamma |
| AI video and avatars | Make clips from text and images with Google Flow and Veo, and make avatar videos with HeyGen, including translation with lip sync |
| APIs and webhooks | Read API documentation, handle authentication and API keys safely, and fix common errors, without writing code |
| No-code automations | Build, debug and schedule a working Make.com scenario with a router and an AI step per branch |
| AI agents | Build, test and publish a customer support chat agent in Voiceflow with a knowledge base and an escalation tool |
| Vibe coding | Plan, generate, test and deploy a simple app with Bolt, Lovable or Google AI Studio, with a database, hosting and a custom domain |
Other topics from the systems we build every day, such as voice agents and AI in CRMs, can be scoped on request.
How does a training engagement run?
A Hyperloop AI training engagement runs in six steps: discovery, scoping against your syllabus, writing the content and demo businesses, recording or delivering, revisions, and handover. Our first course, about 6 hours of video across 7 lessons, was written and recorded over about 3 weeks.
Discovery
We agree who the learners are, what they should be able to do afterwards, the format and where the content will live.Scope against the syllabus
Your outline or goals mapped to lessons, with topics, tools, lengths and resources per lesson.Content and demo businesses
We write the lessons and build the demo materials learners reuse across lessons.Record or deliver
Screen-recorded as working builds, or delivered live as workshops and team sessions.Revisions
You review the lessons; we make reasonable revisions on request.Handover
Videos, guidelines and resources in the format your platform needs.
Who owns the course?
Who owns a Hyperloop AI course is agreed before work starts. Recorded courses can be made as work for hire, so all rights pass to you on delivery and payment, and we don't reuse the lessons or resources without your consent. New courses for other clients are written from scratch.
What are the limits?
AI training has real limits. Tools change fast, so screens in recordings will drift even though the method stays useful. Learners need their own accounts for the tools taught, and some features need a paid plan. Builds use demo data only. And a work-for-hire course can't be reused without your consent.
- Tools change fast. Features and screens shown in recordings will change. Refreshes can be scoped as new work.
- Third-party tools. Learners need their own accounts, and some features need a paid plan.
- Demo data only. Learners adapt the examples to their own business.
- Content ownership. Work-for-hire courses aren't reused without your consent.
Not the right fit if
- You need academic theory or AI research rather than hands-on use of current tools
- You need content that never dates
- You want lessons built on your real client data in recordings
Frequently asked questions
Questions
Answer
Who owns the course?
Ownership of a Hyperloop AI course is agreed before we start. Recorded courses can be made as work for hire, so all rights pass to you on delivery and payment, and we don't reuse the content without your consent. Other arrangements, such as a licence, can be agreed instead.
Can you follow our syllabus?
Yes. We build to your programme outline, including the topics for each module and any minimum video lengths, and show you how each lesson maps to it. If your outline is rough, the scoping step turns it into lessons with agreed topics, tools and lengths before anything is recorded.
Will the lessons go out of date?
The screens will, because AI tools change month to month. The method doesn't. Lessons focus on prompt structure, knowledge base structure and a build process that work in any tool, so the skills still apply when a tool changes. Refreshes can be scoped as new work when the tools change.
Do learners need technical experience?
No. Technical ideas are explained with everyday analogies, such as an API as a waiter and a webhook as a doorbell, and every build is shown step by step, including fixing errors. Learners need their own accounts for the tools taught, and some features need a paid plan.
Can you train our team on our own tools and processes?
Yes. Team training uses your own tools and your processes as the examples, so your team leaves with assistants and automations they can use straight away. We need access to the tools your team uses and the documents you want used as examples.
Is training or a build the better choice?
If you want your team to build and run AI themselves, start with training. If you want a system working without your team building it, we build it for you in AI development. Many businesses do both: a build for the system that matters most, and training so the team can handle the smaller things.
How long does it take to produce a course?
It depends on the scope: the format, the number of lessons and their length. As a guide, our first course, about 6 hours of video across 7 lessons, was written and recorded over about 3 weeks. The timeline for your course is agreed when it's scoped.