Clear prompt structure
Define the goal, context, requirements and expected output. Practise checking an answer and improving it.
February cohort: in-house marketing and communications professionals
From a work challenge to an AI approach you can test.
Two practical morning sessions for in-house marketing and communications professionals across industries. Learn to give AI clear instructions, build your own reusable agents and check their output in content, data and everyday work tasks.
Everyday work
From an idea to a practical exercise
Learn how to build your own AI agents and reusable workflows, rather than just one campaign. In one practical example, we will give an agent brand context, have it ask questions to clarify the brief, and develop initial campaign ideas, visual sketches, copy options and a channel publishing plan. We will then revise the output based on feedback. We will explore a few examples selected for the group in depth; two sessions will not cover full implementation of every solution or automated publishing.
The February cohort is for professionals who handle marketing and communications within a company. Participants can come from different industries, with shared needs around ideas, content, reporting and recurring tasks. The techniques can also be applied to other parts of your work.
No previous AI experience is required. Before the workshop, you will complete a short survey about your daily work, recurring tasks and experience with AI. This helps me choose examples and exercises relevant to the group.
You will need a computer and an internet connection. We will confirm access to tools and any preparation before the workshop.
Practical programme
Define the goal, context, requirements and expected output. Practise checking an answer and improving it.
Build an agent's instructions, workflow and quality checks. Test it on a specific task and learn principles you can use to create other agents.
Develop ideas, initial visual sketches and copy options from a clear brief. Compare the options and refine them based on feedback.
Summarise documents and tables, look for patterns and missing data, and check conclusions. Separate facts from assumptions.
Explore the basics of interactive HTML reports and a simple campaign landing-page prototype. This is not a course in building or publishing a complete website.
Step by step
Alongside the workshop
Try AI tools on tasks similar to those in your working day.
Learn to structure tasks, build your own AI agents and check their output. You can apply these techniques to other tasks in your work.
Receive a selection of practical approaches and prompts, followed by six weeks of weekly practical email tips and shared answers to the group's questions.
One 45-minute online session to discuss questions that arise as you use what you have learned.
My creative direction work at Radio SWH brings together advertising ideas, content and technical delivery. I am building ApprovalDock, a platform for reviewing and approving images, video, audio and PDFs. At Prestigo Media, I work on TV playout systems and bespoke technical solutions. In the workshops, I bring these perspectives together to help participants break down work tasks, try AI tools and check the results.
Before applying
I will contact you after receiving your application and prepare the invoice manually. The invoice will state the payment deadline; I will confirm your place after payment.
Submit an application or email eduards@cers.lv. We will discuss availability individually.
We will agree the required document format and delivery method before payment.
Use anonymised examples. Do not provide personal or sensitive data, passwords, confidential documents or trade secrets.
Individual consulting, complex integrations and software development are outside the programme. This is not advanced programming or machine learning training.
I will send participation, rescheduling and cancellation terms with the offer before payment. Submitting the form alone does not confirm participation.
17.–18.02.2027. 10.00–14.00
From a work challenge to an AI approach you can test.