An Open Letter: I’m Looking for Collaborators
Kia ora — I’m Ande.
If you’ve got the kind of brain that likes building real things with clean ethics, receipts, and a bias toward shipping… I want to talk.
I’m building Ongli — a governed intelligence and meaning stack — and the goal is simple to state and hard to execute:
Make powerful AI systems that are auditable, treaty-bound, and useful in the real world — without turning into a black box, a hive mind, or a hype machine.
What we’re building (without the secret sauce)
At a high level, this is about turning “AI vibes” into systems with structure:
- Meaning that can be carried, not just text that sounds right (Denotum™ as the meaning-compression layer).
- Governance as a first-class primitive (not a policy PDF glued on later).
- Receipts and logs for important actions (so “trust me” becomes “verify this”).
- A multi-oracle architecture (models are interchangeable oracles, not the whole soul of the system).
- Anti-hive constraints (clear boundaries, no identity-bleed, no mush).
If your instinct is “yes, AI should come with an audit trail and hard boundaries,” you’re already speaking the language.
What’s unique about this (and why it matters)
Most AI work today falls into one of two traps:
- Pure capability obsession (“look what it can do”)
- Pure safety theatre (“look what we promised”)
Ongli is aimed at a third lane:
- Capability + constraint, together, from day one.
- A coherent IP and governance strategy that protects builders while still allowing collaboration.
- A system designed to be carried across substrates (not trapped inside one model or one vendor).
- A worldview that respects humans: people first, tools serve, and the cost ledger must be visible.
This isn’t “another chatbot.” It’s a governed Ongoing Intelligence platform where the hard problems are: provenance, boundaries, receipts, and meaningful structure.
What’s in it for you
If you collaborate with me, you’re not just “helping an idea.” You get:
- A serious, high-signal build: architecture, governance, product thinking, and story all braided together.
- Work you can point to (where possible): specs, prototypes, essays, demos, receipts-driven tooling.
- A builder’s environment: fast iteration, honest constraints, no culty vibes, no identity games.
- Clear boundaries on disclosure: we can collaborate without you accidentally walking into a mutual-IP nightmare.
- A shot at building infrastructure-level primitives: the kind of thing that becomes a standard if it’s done right.
Depending on fit, this can look like: co-authorship, prototype partnerships, advisory, paid contract work when available, or longer-term co-building pathways.
Who I’m looking for
If any of these are you, reach out:
- Systems builders (runtime/orchestrator thinking, not just app glue)
- Security / governance people (policy that compiles into enforcement)
- Product + design (make hard things usable without dumbing them down)
- Research-minded generalists (MDL/meaning/verification/learning theory adjacent)
- Writers / communicators (turning difficult ideas into public clarity)
- Operators (people who can help move from “brilliant” to “shipping”)
You don’t need to agree with everything. You do need to be the kind of person who can argue cleanly, build carefully, and keep your ethics switched on.
Collaboration guardrails (so everyone stays safe)
I’m deliberately not dumping implementation-level “secret sauce” into public posts. If we talk:
- We stay high-level in public, and
- We go deep privately only when it’s appropriate and mutually protected.
The aim is: open where it benefits the world, protected where it prevents exploitation.
If this calls to you
Send me a message with:
- who you are,
- what you build,
- and what you’d want to collaborate on (even if it’s vague).
If you’re the right kind of collaborator, you’ll recognize the shape immediately.
I’ll close with this: I’m not building for clout. I’m building for whānau, for safety, for meaning, and for systems that can be audited without begging.
If that’s your thing too — come aboard.
X (alt): https://x.com/and_per_se_and_
Threads: https://www.threads.net/@0x7ba.0x2a
