The Challenge: Moving a Cross-Functional Team Beyond Foundational AI Use
By the time of the second run, most participants already had access to Generative AI tools through their day to day work. The gap was not awareness, it was operationalisation: teams knew the tools existed but had not worked out how to weave them into their actual workstreams in a way that stuck.
Sembcorp needed a programme that could take participants past casual, ad hoc use and toward a genuine operating model for AI in their roles. At the same time, the training had to deliver groundbreaking takeaways, insights and techniques that felt new and relevant even to a group that had already been introduced to Generative AI, rather than a repeat of ground they had covered before.
Why Heicoders Academy
Sembcorp returned to Heicoders Academy because the first engagement had already proven the model: a curriculum customised to their teams, taught by instructors who work hands on in tech rather than delivering theory from a slide deck. That track record made the decision to commission a second run straightforward.
Repeating the partnership also let our instructors build on what worked in the first cohort, refining the curriculum with sharper use cases and more ambitious project briefs for participants who came in with a stronger baseline understanding of Generative AI.
The Solution
The second run of the customised GA100 programme continued to combine prompt engineering fundamentals with hands on work building AI agents and workflow based tools. Participants worked through:
- Prompt Engineering: Refining techniques for getting consistent, high quality output from large language models.
- AI Agents and Automation: Building and deploying agents, including tools like Manus, to handle multi step tasks.
- Workflow Based Agent Design: Structuring AI agents around real business processes rather than one off prompts.
- Generative AI Toolkit: Working across a broader creative and productivity toolkit to extend what participants could build.
- Applied Capstone Projects: Turning the above into deployed, working applications rather than slide based proposals.
The programme was led by Kong Yu Ning, our Chief of Growth and Principal Instructor.
Two of the standout capstone projects from this run were a 3D furniture modeller, a shop front that lets customers visualise furniture pieces in their own space before purchase, and an HSSE Inspection and Audit tool built to support health, safety, security and environmental inspection workflows within Sembcorp's operations. Both moved well past a proof of concept prompt and into deployed, usable applications, a clear sign of how far the cohort's applied AI skills had come by the end of the workshop.
Results
The second cohort left the workshop having built and deployed real tools rather than one off demos, a meaningful step up from prompting exercises alone. The 3D modeller and the HSSE inspection tool in particular showed that participants could take a genuine operational need, whether in customer facing product visualisation or in safety and compliance workflows, and turn it into a working application within the scope of the programme.
What's Next
Building on two successful runs, Heicoders Academy and Sembcorp Industries are continuing the partnership, with further sessions planned to extend AI ready capability to more teams across the organisation.
