HOW IT NARROWED
Mesh started as something else. It was a dashboard, then it was three products wearing one name, and then I took most of it out.
This is how it got from there to one platform doing one job for training providers. The mess is left in, because the mess is the reason it works now.
I spent years buying software that solved most of a problem and left the rest to me. The first version of this was built on that instinct, and it was modest: take the numbers a business already had, spread across the systems it already paid for, and put them on one screen.
Then it grew. A CRM went in. So did HR and staff wellbeing, and business intelligence to make sense of the rest.
It looked like one platform. It was three, and it hinted at a lot of problems without solving any of them well.
Every customer wanted something different, and I built for all of them. One wanted a wellbeing snapshot of his staff. Another had no interest in that and wanted the sales side instead. Both were reasonable, and saying yes to both is how you end up with three platforms running in parallel, each about eighty percent finished.
Then something would pause. Money moved, attention moved, and some of those pauses ran eighteen months. A new biggest problem would arrive and we'd start again. Years of it — and a client with a website and a booking system to show for the lot.
One afternoon in November a client rang to ask how far away the customer portal was — the part where the people booking his courses could manage their own enrolments.
It wasn't in the system I was building. It was in the one before it.
That was the diagnosis, and it arrived as a question. Everything went into one application over Christmas. His portal was live by mid-January.
A month later he rang about a new customer and mentioned, in passing, that she'd only signed up because of the portal. The part built almost as an afterthought was the part people were choosing him for.
And the ones who needed it weren't the students. It was the bookers — the people accountable for keeping a workplace compliant, who need to see who's certified, who's expiring, and who to put on the next course.
I took two things out.
Business intelligence got parked. It's a finishing touch, not a leading edge, and it was never the reason anyone showed up.
HR and wellbeing came out whole and went somewhere else. Not because the problem wasn't real — it was, and it still is — but because I couldn't solve it properly inside a platform that was already trying to be two other things.
What was left was one platform, one kind of business, one job.
What happened to the wellbeing work →SI-Brain, the intelligence underneath SI-Ops, became the foundation of Coach Core. Instead of a feature called AI, each product gets a named companion. Solvent's is Eli. Mesh's Sales Coach and Training Coach are planned, not built.
Business intelligence comes back too, somewhere it makes sense: as Coach Core is implemented, it trains each companion on the business's own policies, procedures and performance.
Mesh is course management and a student LMS in one platform, and it's onboarding training providers now.
One provider kept their schedule current by emailing support, because there was no way to import it. Flexible course templates and ticket sales mean a provider can sell what they actually run.