Corporate AI training in Hong Kong: choose what your team will build

A live AI demonstration tells you what the trainer can do. A useful workshop gives you time to find out what your team can do without the trainer touching the keyboard.

Choose an outcome your team can attempt during the session.
Match the format to the people attending: a company workshop and an assessed course serve different needs.
Ask what happens when participants get stuck, both in the room and afterwards.

I used the providers’ published formats for this comparison. Confirm the current syllabus with each provider; its standard programmes and a custom company session can differ.

Compare the format first

Swipe across the table to compare all columns.

Provider	What its current pages describe	Consider it when	Clarify before booking
Lyfar Studio	One-to-one lessons, small-group masterclasses, hands-on workshops and custom company sessions covering research, knowledge systems and automation. Education offer	A small team wants direct guidance from a builder on a defined workflow.	Your prerequisites, the system participants will attempt and the support included in your booking.
Preface	Tailored corporate technology training, including lunch-and-learn sessions, workshops and bootcamps; published topics include AI agents and generative AI prompting. Professional services	You want to discuss several training formats or a broader company programme.	The assigned instructor, participant grouping and time spent practising.
HKPC Academy	Custom in-company training delivered in physical, online or blended formats. Its FutureSkills framework includes AI workflows, retrieval and agentic AI. Corporate training, FutureSkills	Your training owner wants to compare a tailored programme across defined skills.	Which parts of the framework appear in the actual session proposal.
HKU SPACE	Its published Office Administration module covers generative AI alongside office applications and includes assessment. Programme page	An employee needs structured study with an assessed outcome.	The next available intake and entry requirements. This course listing doesn’t establish a bespoke company workshop.

The providers above overlap. Ask to see the exercise plan, including the time participants spend building and reviewing their own work.

The session I can show you

My March 2026 masterclass at HKU covered AI-assisted research with source documents, retrieval and live demonstrations. The published recap shows the group and explains the workflow from that session.

One part of that recap still matters when I plan training: I described a plausible formula the model had given me that didn’t survive checking against the source. I use that failure to explain why participants need to inspect evidence. A fluent answer can consume the whole room’s attention before anyone checks it.

The knowledge-management system is another example I can bring into a session. It is a system I built for my own work, with connected notes and retrieval. It gives us something concrete to examine, then adapt to the team’s problem.

A corporate workshop needs its own brief. Your sales team might need different tools from a research group. Agree whether you want participants to learn a method or need someone to deliver a production system; those require different scopes.

Write the learning outcome as a task

“Understand AI” gives you little to check afterwards. Try a more precise request: participants should assemble a research note from a set of approved documents and show the source behind each material claim. Or they should build a draft workflow that routes an enquiry to the right person and asks for missing information.

Choose a task that resembles the team’s work and fits the available time.

Ask the provider to reserve time for participants to attempt it. A cooking class where only the instructor uses the stove leaves you with a good lunch and an uncertain dinner plan. You need time to make mistakes with someone nearby who can explain them.

Send the trainer the group’s starting level and approved tools before the session. Decide whether attendees can install software, create accounts or use company information. Use prepared sample material when permission to use real documents is unclear.

Agree the follow-through

Ask who receives the exercise files and whether attendees can continue using the tools afterwards. Confirm any subscription requirements. Establish what support means: written notes, a follow-up call or help debugging the exercise are different commitments.

For a Hong Kong team, specify the teaching language and the languages used in the work itself. Include mixed-language examples if they are relevant to the job. Ask the trainer to demonstrate the exercise with that material before settling the scope.

For a team that needs a system delivered and maintained, use the AI-automation provider comparison to assess that separate purchase.

If you want a small team to work on a defined research or automation task with me, send the task and the group’s starting level . We can then decide whether a lesson, workshop or a separate implementation project will serve you.