
Beyond the Hype: How We Spent the Last Few Months Actually Building Agentic AI
Everyone is talking about AI these days. Between sci‑fi predictions, breathless keynote speeches, and endless LinkedIn think pieces, it is easy to forget one thing: none of that matters unless the technology actually does something.
So over the past few months, we stopped talking and started building.
We have been deep in the trenches – wiring systems together, testing flows, breaking things, fixing them, and turning buzzwords into working capabilities.
What follows is a look at several AI‑powered solutions we have recently deployed. They are real, they are running, and they are already changing how people work.
Agentic AI - The Brain and The Hands
In a typical Agentic AI setup, the agent is the brain: it interprets what the user wants, decides the next step, and orchestrates the work. The scenario or flow is the hands – the structured, deterministic sequence that actually performs the tasks.
The agent thinks, reasons, and chooses; the flow executes, retrieves, updates, and returns results. Together, they form a small AI “team” where intelligence and action reinforce each other.
The SOP Whisperer - Agentic AI Driving Chatbots
Chatbots are everywhere. Most of them are glorified FAQ machines. Ours is different.
We deployed an AI agent directly on our website that serves two audiences at once:
Visitors, who get instant answers to general questions
Colleagues, who get precise, internal process guidance
The magic lies behind the scenes. We attached a curated body of organisational knowledge to the agent, turning it into a real‑time oracle for process questions.
For example, our CHL process simulation experiences a boost by this agent. “What is the location for CHL customer Lean Mei” is immediately answered with “The location for Lean Mei is East Coast” (Make.com)
No hunting through SOPs. No waiting for someone to reply. Just answers – instantly.
This agent is built on Make.com, orchestrating webhooks, knowledge modules, and email responses in a clean, automated flow. It’s already boosting our CHL process simulation work and reducing friction across the board.
The Executive Sleep-Saver - Agentic AI Never Sleeps
As I recently highlighted, AI is at its best when it gives people their time – and sanity – back.
Case in point: we built an automated news-compilation agent that gathers, synthesises, and formats essential briefings. We deployed this for a government officer who previously spent hours manually reading and compiling updates for his minister to be on his desk at 09:00h every morning. Now, the agent handles the heavy lifting.
The minister gets a crisp, highly accurate brief right on schedule, and the officer gets to sleep a little longer in the morning.
The minister still gets a sharp, accurate briefing at 09:00. The officer gets something even better: more sleep.
Here’s a sample of what the agent produces — clean, structured, and ready for action
This isn’t hype. It is practical agentic AI that removes cognitive load and frees up human attention for real work.
The Magic Glue: Agentic AI Powered by Make.com, Copilot and PAD
AI without a delivery mechanism is just a very smart calculator. To make it truly useful, you need the right plumbing. We’ve been busy architecting complex, multi-step automated flows using Make.com and Power Automate Desktop (PAD).
Think of these flows as tireless digital colleagues:
They never sleep
They never complain about scope creep
They never forget a step and
They never miss a detail
(Though, admittedly, they still can’t fetch kopi and toast from the hawker centre. We are working on it.)
These flows are the backbone of our agentic systems – connecting knowledge bases, orchestrating data, and ensuring that AI outputs actually reach the people who need them.
From "AI Curious" to "AI Capable"
Having a strategy is great, but execution is where the magic happens.
Whether it is:
deploying smart agents that handle daily operational grind
integrating AI into public administration
aligning automation with security and workforce frameworks or
building agentic architectures that scale
– we’ve gained the technical battle scars to make AI work in real environments.
Our Change Management Navigator helps teams move from high‑level ambition to on‑the‑ground adoption. We bridge the gap between leadership vision and technical reality.
We bridge the gap between high-level strategic management and on-the-ground technical execution. If you are ready to stop talking about AI and start putting it to work for your team, we should talk.
Conclusion: Agentic AI - Our New Support Staff Has Arrived
The era of “AI is a novelty” is over.
The era of “AI is a utility” has begun.
We have spent the last few months proving that with the right architecture, the right flows, and a bit of creativity, you can completely transform how work gets done.
We build the systems so you can get back to the strategy. Ready to see what an automated workforce can do for your bottom line? Let’s talk.
Can we build your next automated solution?
Can.
- All Posts
- AI
- BPR
- Career Development
- Cases
- Change
- Coaching
- Competency
- Customers
- Data Science
- Enablers
- HR Strategy
- Innovation
- Leadership
- Lean Six Sigma
- Manufacturing
- OD
- Operations
- Service
- Staff Development
- Strategy
- Workforce Planning
- Working Capital
- YOG
- Back
- DISC











