Påmin Labs — questions and answers

Plain answers about the studio and the two things it builds. Everything here is also in llms.txt for machine readers.

1. What is Påmin Labs?

   Påmin Labs is an applied research studio based in Sydney, Australia. It builds infrastructure for AI that has to keep working over time — the parts a person takes for granted at work but that are missing beneath today's agents: memory, delegation, review and boundaries. Everything it ships is open source under MIT, and every claim on its pages points at a number the repository can reproduce.

2. What does Påmin Labs build?

   Two things. Påmin Memory is a memory layer for AI agents that keeps the source trail intact — versioned facts, explainable retrieval, local-first, written in Rust. Orchestrator is an AI team for a company of one: you say one sentence, approve a plan and merge, while ten roles build, test and review inside containers you never open. Both are separate products with separate repositories.

3. Is Påmin Labs' work open source?

   Yes. Both products are public on GitHub under the MIT licence, at github.com/pamin-labs/PaminMemory and github.com/pamin-labs/orchestrator. The roadmap on each product site is read against the real date rather than hand-edited, so a milestone stops claiming to be in progress once its window has passed.

4. How are Påmin Memory and Orchestrator related?

   They answer two halves of the same problem. Memory is about what an agent knows across time; Orchestrator is about how several agents divide work without a human holding every thread. They are usable independently — Orchestrator does not require Påmin Memory, and Påmin Memory plugs into any agent, not only Orchestrator.

5. What research questions is the studio working on?

   Four: how an agent should hold a fact that used to be true (temporal knowledge); whether retrieval can explain why it chose what it chose, from the same evidence (explainable retrieval); where trust lives when ten agents share one repository (delegation with boundaries); and what "done" means when the author is a model (measured autonomy).

6. Who is behind Påmin Labs?

   Three founders. Jason — systems, architecture and product, including the low-level implementation. Josh — product tuning: how the tools feel in someone's hands and what to leave out. Arvin — machine learning: retrieval, ranking and the models behind memory. The team is small on purpose.

7. How do I get in touch?

   Email info@paminlabs.com for research partnerships, early integrations or a question. Issues and pull requests on either GitHub repository are read too.