Getting to know Tiffany Ong
Meet Tiffany Ong, Head of Microsoft Singapore at Wild Tech, as she shares her career journey, leadership approach, customer focus and what makes Wild Tech’s culture unique.
Blog
I’ve been the Microsoft Business Central champion within Wild Tech for a few years, and we’re constantly building Copilot Agents on top of it.
Since I sit in sales myself, I want to share one that is close to home: a sales agent built to speed up quoting for mid-market Australian manufacturers and supply chain businesses.
The agent’s objective:
In this demonstration, I ran through how the agent runs the same quote request through three different starting points: a voice note after a client meeting, typing into my phone’s Copilot app and from a website enquiry.
In sales, you’re constantly on the road running from meeting to meeting so that the agent can work from wherever you are.
One caveat before I get into it. None of this works unless Copilot has data it can actually fetch, and trust once it’s fetched it. That means item costs, stock levels, customer terms and credit status all sitting somewhere structured and accurate. Business Central makes that part easy, because it’s already the single source of truth for all of it. If that data is scattered across separate systems that don’t talk to each other, this doesn’t work, no matter how good the AI is.
Now that we’ve got that out of the way, let’s go through each Starting Point.
“Hi there, I’ve just come out of a meeting with School of Fine Art and they’ve asked if we can have around 50 kilograms of mango pulp delivered by the 30th of September.”
That’s the dictation I gave the agent directly in the CRM as you can see in the screengrab below.
I stopped to check it was accurate, tidied it up to “50 kilograms of mango pulp” so the system could work out the unit of measure itself using the ERP, and then said: “Create me the most profitable quote”.
From there, the agent goes and works in the background, looking into Business Central, looking at the items, and working out what we actually have on stock and on hand. It creates 3 separate quote options before running a side-by-side comparison.
In order to prepare these quotes, the agent is pulling data out of Business Central to evaluate the:
For example – is there a location that we have stock available where it actually could be a bit cheaper for us to send that stock to that customer from that location instead of somewhere else?
So there’s all these different components that would usually take us hours or days to review but can now be done in minutes.
In the screengrab below you can see it’s pulling out Unit Pricing and identifying potential risks with the quote and critically, calculating the gross profit and margin for this quote option.
Once the agent has developed three options for review, it develops a helpful side-by-side comparison in order to recommend the most profitable, risk-adjusted option for you to use as a quote.
In some cases, working out the stock on hand, the inventory availability, the locations, and the most efficient and cheapest way to fulfil an order can take a couple of days for some of the clients we work with. Within that one minute, the agent had gone into Business Central, pulled up the customer account and contact, and checked any credit holds or terms we needed to be aware of.
We had the request, the lead time, all of it, and then it worked out the most profitable way to quote that customer while keeping us protected.
There were three quotes on the table. Option A: a custom request on a single shipment. Option B: a slight premium. Option C: standing list price. The agent ran a complete side-by-side comparison of all three. Option A came in at a 38% margin. Option B was a bit of a stretch, but landed a 41.7% gross margin, and it flagged the fulfilment risk to the customer for each option as well.
What the agent told us was that the most profitable option was actually B, the highest gross profit, about $200 more than option A, with the highest margin, and it gave its reasoning: three independent premium drivers similar to option A, but a tight deadline attached. Even so, the recommended option was A, not B, because of that risk.
That’s the part I want to call out. We didn’t have to take option A. But now, as a sales rep, I had all of that information in front of me to make a genuinely informed decision, one that gets speed back to the client, protects the margin the business needs given rising costs, and accounts for whether the order is actually feasible to fulfil.
I was happy with the recommendation, so I told the agent: create option A, and draft an email to the client. That’s it. That’s the whole decision, made with better information than I’d normally have, in about a minute.
Now to circle back on the two other starting points I mentioned.
The example I’ve just given you was when I was in front of my laptop, within the Business Central CRM, instructing it with a voice dictation.
But I didn’t really need to be at a laptop to drive that result.
I can do the same thing from my phone: pull up the same agent on mobile and say, can you create me a quote in Business Central.
Where a prospect gets in touch via your website form, social accounts or directly via email – it’s the same outcome.
This agent is a live example of what we’re building right now inside Business Central for manufacturing and supply chain sales teams.
The AI didn’t get smarter on the day. It worked because the data underneath it was already structured well enough to use. That’s the real question for most businesses watching this space: not whether Copilot can do this, that part is proven, but whether your systems are ready for it to try.
For eligible businesses, we’re offering a 2-hour workshop to review your systems and see if this is something you can start to use for your team.
If that’s of interest, let us know by getting in touch.
Contact Us
Meet Tiffany Ong, Head of Microsoft Singapore at Wild Tech, as she shares her career journey, leadership approach, customer focus and what makes Wild Tech’s culture unique.
Meet Louwis Venter, Wild Tech’s Executive GM of AI Innovation, as he shares insights on AI, career growth and building a culture of creativity, ownership and impact.
NetSuite in 2026 moves toward AI-native ERP. From agentic workflows to MCP connectivity, here’s how to prepare your instance and governance.
We’re here to listen and to solve your problems then help you to digitalise your business. Let’s talk.