What building real tools with AI looks like in 2026.
In 2026, the most useful signal of AI competence is simple: what have you actually shipped?
Not what you’ve talked about. Not what you’ve advised others to do. What have you built, released, and made available for real people to use?
I’ve been answering that question the only way that matters — by building. Over the past year I’ve been creating internal tools for myself, for Thel Consulting, and for clients. Some stay private. Some become shared resources. All of them exist because they solved a real need faster than traditional approaches would have allowed.
A Recent Example
One of the more public ones is the South Dakota Technology Commercialization Resources map.
I needed a clean, searchable way to surface government and public university programs that support technology commercialization across the state — something founders, researchers, and economic development staff could actually use. From the first prompt to a live, embeddable widget took about thirty minutes.
The result is now available on our About page. It maps public resources by city, includes search and filters (funding, SBIR, defense, and more), and has a “Suggest Edit” button so the community can keep the information current.
It is a working tool, not a polished product (there's some clear errors in this initial version). The data will need ongoing fact-checking — programs change, funding levels shift. That’s why the edit button is there. If you see something missing or outdated, use it.
Why This Matters
The point is not the speed, though thirty minutes from idea to live tool is worth noting. The point is that useful tools can now be built and released at the pace of actual need.
I do this regularly:
- Internal tools that make my own work faster and more consistent
- Client-facing resources that solve specific problems in the innovation and commercialization space
- Public demos like this map when they create value for the broader South Dakota ecosystem
When someone positions themselves around AI expertise, the most productive question is straightforward: What have you shipped that other people can actually use?
The rest is conversation. Shipping is the work.
