In early 2026, we hosted an AI Workflow Automation workshop for a group of operational leaders from the Shopee team. A pre-workshop survey revealed something we didn't expect: while most participants used AI daily, they were heavily bottlenecked by "data paralysis." The single most cited execution blocker was waiting for BI data exports, and many leaders reported losing 5 to 10 hours a week on repetitive administrative tasks. When searching for internal policies, they relied on tribal knowledge — pinging colleagues in chat groups or manually digging through Drive folders.
To address this, we split the training into two phases designed to shift their mindset from "copy-pasters" to workflow builders.
Phase 1: The Data-First Principle
Workshop 1 focused on the Data-First Principle — teaching the leaders that AI is only as good as the context you give it. We demonstrated how to build a centralised "Second Brain" using NotebookLM, consolidating scattered SOPs and project trackers into an interactive, shared knowledge base. The goal: eliminate the reliance on manual chat searches and make institutional knowledge actually searchable.
We also showed them how to bypass BI wait times by using AI to reverse-engineer their own complex trackers and generate executive summaries on the spot. The look on their faces when a 3-day BI request was replicated in 10 minutes was worth the entire workshop.
Phase 2: From Systems to Automation
Workshop 2 brought these systems into the marketing funnel. We tackled rapid prototyping — showing the team how to use tools like Google AI Studio and Lovable.dev to deploy website prototypes in minutes, skipping the traditional design → spec → dev cycle entirely. Then we moved into no-code CRM automation, connecting Google Forms to Node-RED and Telegram to instantly classify leads and notify sales teams.
What Surprised Us
By the end of Workshop 2, these leaders had gone from copying ChatGPT outputs into spreadsheets to building automated pipelines that classify leads with zero manual effort.
But the thing that stuck with us most wasn't the technical outcome. It was how quickly the participants' questions changed. In Phase 1, they asked "How do I prompt better?" By Phase 2, they were asking "Can I connect this to our internal system?" That shift — from prompting to building — is the whole point.
If we ran this workshop again, we'd spend even more time on Phase 1. The data organisation step felt tedious to some participants at first, but every single person referenced it as the most valuable part in their feedback forms. Turns out, the boring foundation work is what makes the exciting automation work actually stick.
We share our workshop frameworks and materials on the Alpha Bits blog. If you're building something similar for your own team, our Data-First Principle Thinking post covers the framework in detail.