Josiah Forestieri

I was hired in February 2020 to cold-call restaurants, and started the month the country shut them.

I have been at FoodByUs ever since, through five job titles, and the job has changed a lot more than the titles say. It started with a phone and a list of venues. Now I build the software the sales and support teams run on: find where a team is losing its week, size the problem, work out what the tool should actually do, and get it into their hands.

The job was always the phone: ring hospitality venues and book meetings for the sales reps. I did it from my bedroom. Cold-calling hospitality in the year hospitality shut is not an intro to sales anyone would design, and I learnt more from it than I would have from a good year.

By 2021 I was managing the function I had been the only person in, and grew it to five reps across New South Wales, Victoria and Queensland. I wrote the training plan and the call scripts, ran the hiring, and put in the tools the team worked out of. A few of the people I trained were promoted into full sales roles, which is still the part I am most pleased about.

I moved into the operations team and picked up a second team doing onboarding and support work while still running sales development. It was my call rather than a restructure. I had been doing the operational half of my own job for a while, building out HubSpot, and I liked that half more.

I arrived with nothing technical worth the name. Spreadsheets, reporting and workflow tools were where I started, and everything below comes out of that. The two years after it were process work: build the way a task gets done, test it, find out where it does not hold, fix it. In November 2024 the support function was restructured and I moved into a programs role in ops.

These days ops means software. I build what the reps and the support team actually work out of: flows that pull buyer and supplier details off scanned invoices, a notification that tells a rep when one of their accounts is about to spend somewhere else, and the price comparison tool below. I also moved our support platform off Zendesk and onto HubSpot, five channels and eight years of history.

A rep wins a wholesale account by showing a venue what they are overpaying: every line of their invoices matched against our catalogue and totalled. A rep used to request one from the support team and wait three days or more to get it back.

My first version automated the matching and still walked the rep through six screens of it. So I changed what the product was for: the rep does not want to co-pilot a price comparison, the rep wants the price comparison. A rep gets one back in about thirty minutes now, it is live and they are using it, and it is still not as accurate as the hand-built version it replaced. That last part is the work that is left.

The move that keeps coming up is checking the thing everyone assumes is already true. There was exactly one event in our analytics built to catch a buyer about to spend somewhere else, and when I went to use it I found it had never fired in production, not once. The vendor features that were going to make the support migration straightforward worked in the sales deck and not in our account.

I run my own projects like a small engineering org and the labour is AI agents. Every project has a file that says where to look. Bigger work runs as a map of numbered tickets, each one resolving a single decision. Settled decisions go on a do-not-reopen list, because a fresh session has no memory of the argument that closed something. Nothing ships until a reviewer with no memory of it reproduces the number.

None of this started with AI. The expensive part was never producing the answer, it is having to produce it twice.

I have been teaching the thing I do for about as long as I have been doing it. My own team at ING Direct six months in, three years as a teaching assistant at a school, then the sales team I had built, then onboarding, then support. Lately it is AI, one on one with our CRO.

A way of working only I can run is worth whatever I can personally produce. One that other people pick up is worth a fair bit more.

Every CV has a what-do-you-do-in-your-own-time question somewhere on it. The honest answer for me is far too much. I like learning new things, and this is what that adds up to: everything I have picked up since 1999, when it started, and whether it stuck.

Some lasted two decades and some lasted a year, and the chart does not hide which is which. Same instinct as the job: take the thing apart and find out how it is put together.