Practical AI systems for small businesses — AI assistants that live on a website and answer questions and capture leads, missed-call auto-text-back for busy trades and service businesses, and workflow automation for the manual admin that eats a business owner's week. Nothing exotic. The goal is always the same: something that gets used, not something that gets demoed once and shelved.

See the bottleneck → scope the fix precisely → direct AI to build it → make sure people actually use it.

That last step is the one most people skip. I've watched good tools die because nobody was accountable for adoption — so I treat rollout and coaching as part of the build, not an afterthought.

I currently manage 20+ foreign teachers at Vinalearn in Ho Chi Minh City. When manual observation reporting was eating hours every week, I didn't wait for a system to be handed to me — I built one myself, then ran the onboarding and coaching to get a team that hadn't asked for a new tool to actually use it. That's the same job I do now for business owners: find what's actually broken, build the fix, make sure it sticks.

• A manual process that quietly eats hours every week
• A team, or a business owner, who's skeptical of new tools until they see one that actually works
• Leads or enquiries that go cold because nobody responded fast enough
• A workflow nobody has ever properly written down, let alone automated

More of this: fewer slide decks, more shipped tools that a real team or a real customer actually touches. If that's the kind of problem you have, I'd like to hear about it.