AI Knowledge Session with the Rotary Club of Bangkok: If AI Is Going to Work, It Has to Start with Data

TH

AI is being talked about more than ever. But one question still comes up again and again.

“How do you actually use AI in day-to-day work, and trust the results it gives you?”

On January 29, 2026, the Rotary Club of Bangkok hosted an AI Knowledge Session, creating space for leaders from different industries to share perspectives on how AI is really being used inside organizations.

During the session, Aubin Samacoits, Director of Data at Sertis, shared lessons from real projects alongside a broader view of how AI is evolving. One idea came through clearly: AI that works in practice doesn’t begin with the newest model. It begins with something much more basic and often overlooked.

The Key Point: AI That Delivers Results Starts with Data, Not Models

A common pattern shows up in many organizations. 

AI initiatives often begin with excitement around tools or advanced models. But once teams start working with them, familiar problems quickly surface.

Data lives in different systems. Teams use different definitions. There’s no single source everyone truly trusts.

When that happens, AI outputs feel inconsistent. They’re hard to validate, and teams hesitate to rely on them for real decisions.

If AI is meant to support real work, the foundation has to be in place first: Data that’s ready to use. Information that’s connected to a shared view. Quality that people can trust.

Without that, even the most advanced AI struggles to deliver value.

Using AI in Practice: Start Where the Work Hurts

What made the session especially concrete was the discussion of real use cases, not to showcase technology, but to show where AI can genuinely help everyday work.

Examples came from across industries:

• Manufacturing: Improving quality checks, reducing waste, and cutting rework

• Logistics: Planning routes and schedules more efficiently

• Real estate: Finding information faster and communicating more accurately with customers

• Healthcare: Supporting X-ray analysis to ease workload and spot risks earlier

• Banking: Helping staff access information quickly, reduce errors, and serve customers faster

Across all of them, the role of AI was the same: to reduce repetitive tasks, improve accuracy, and help people make better decisions, not replace them.

What Comes Next: From Answering Questions to Supporting Work

Another topic that sparked interest was Agentic AI, systems that don’t just respond, but help plan, coordinate, and take action.

But getting there requires more than capability. 

Organizations need trusted data and clear rules around how AI is used so it can operate safely, transparently, and at scale.

If You’re Starting with AI, Ask These 3 Questions First

To avoid AI projects stopping at pilots or demos, the session closed with three practical questions worth answering early:

• What part of the work do we actually want AI to help with?

• Where does the data live, and how reliable is it?

• How will teams use it consistently, securely, and with confidence?

  
  

If your organization is exploring practical ways to apply AI, or looking into Knowledge Management AI or Enterprise AI to help teams find information faster and make clearer decisions,

Sertis team is happy to continue the conversation. Get in touch here: https://bit.ly/3O5wxq9