Resources · Library
Whitepapers
Deep dives into emerging technologies, AI systems, infrastructure and real-world applications.
Building AI Factories That Deliver
Why writing software got cheap and owning it did not, and what a line that ships internal tools to production needs: a foundation, a floor and an approval.
AI factories 02The App Is the Easy Part. The Foundation Is Not.
Documented incidents where apps built from scratch or generated from a prompt leaked data, overreached, stopped scaling, went unmaintained or were reached through their dependencies.
Foundations 03Build It from Open Source? Count the Seams First.
What Curiosity Studio ships out of the box, what an assembled open source stack leaves your team to build, and why that matters for apps you describe and ship.
Build or buy 04What a Search Score Means
What the number next to a search result is, how it is made, and why it is not a percentage.
Search scores 05How Curiosity Processes Your Data
The steps data goes through in a Curiosity Studio system, from the connector that loads it to the interface your users see.
Data processing 06What You Need Before You Start
The hardware, network, data, language model, people and permissions a Curiosity Studio project requires, and what a pilot can skip.
Setup requirements 07An AI Admin You Can Approve
How Sudo, the admin assistant in Curiosity Studio, changes a production configuration without holding the authority to do harm.
Change control 08The Vendor Lock-In You Always Wish You Had
Every AI platform locks you into something. That is a trade-off. Five layers where your work builds up, how Curiosity handles each, and how to test a vendor’s exit.
Platform choice 09Inside Sudo: A Sandbox, a Diff, a Person
The sandbox, its limits and the approval mechanics behind Sudo, for platform admins and security architects.
Sandbox and approval 10Working with Sudo
What to ask Sudo, how a request becomes one diff, how to review it, and how to teach the assistant.
Admin assistant 11Modern Enterprise Search
The six layers of search, keyword, fuzzy, vector, hybrid, graph and agents: how each works, and where each belongs.
Search architectureYour data. Your infrastructure.