We build software that reasons, not software that guesses.

Noesis: AI you can put under contract.

A neurosymbolic runtime that answers only what it can prove from a governed memory, cites every premise, quarantines what it merely believes — and says so, out loud, when it doesn't know.

No black boxes: every answer is proven, cited, a labeled belief, or a loud abstention — never a guess dressed as a fact.

0
false proofs across 100K synthetic patients and 3.1M derivations — a scoped, gated result held as a standing regression check.
18
reasoners, one provability contract — abstention is first-class; a heuristic is never laundered into a proof.
JNC-8
proven-equivalent to hand-coded guideline logic on the tested patient grid — proven, not asserted.

Every answer wears its mode. Here is one in Proven — open it and check the chain yourself.

Does fluconazole interact with warfarin?
Proven

Yes — fluconazole inhibits CYP2C9, the enzyme that clears warfarin, so warfarin exposure rises and bleeding risk increases. Every step below cites an FDA label.

Open the premise chain
  1. Fluconazole inhibits CYP2C9 FDA label — fluconazole
  2. Warfarin is cleared by CYP2C9 FDA label — warfarin
  3. Inhibiting the clearing enzyme raises warfarin exposure interaction rule
  4. Therefore: elevated bleeding risk — monitor INR derived · proven

An illustration of the Proven mode. When Noesis can't ground a chain, it abstains out loud instead — refusal is a feature, not a failure.

See how Noesis works Join the pilot waitlist

How we plan to make money: the citations above are not only a safety feature — they are an economic one. A system that can prove which source it used is a system that can pay that source. We sell trustworthy capability today, and we are building toward a model where knowledge owners are attributed and compensated when their knowledge is used — instead of scraped, dissolved into weights, or licensed once and consumed forever. Read the business model →

Research Notes

Research Notes · Aug 27, 2026

How the Work Actually Flows

How the Work Actually Flows Most writing about AI agents shows you the last ten minutes: someone types a request, an agent writes code, the code works. That is the

Research Notes · Aug 27, 2026

Many Agents, One Rule

lands on feature-b main's log -> does not contain it A's own git log -> looks correct to A nothing errored, nothing warned With a worktree: A stays on main,

Research Notes · Aug 27, 2026

Skills: Stop Retyping Instructions

Error: No such option '--board' (create --help: 0 mentions of board) board-add research --preset hdd -> rc 0, installs NO hdd skills; /hypothesis /experiment /idea /literature all absent, all four

Research Notes · Aug 27, 2026

Ten Minutes to a Board

arrow-kanban 43s cold, rustc 1.97.1, macOS arm64 init --theme software 0s, .arrow-kanban/ + 7 templates create feature … 0s, "Created FT-1300" board / list render as shown items.parquet 6140 bytes,

Research Notes · Aug 27, 2026

When a Tool Earns Its Place

When a Tool Earns Its Place You have run the loop a few times now. Each hypothesis is a file, each result is a file next to it, and every

Research Notes · Aug 27, 2026

AI Self-Improvement, and Why It Is All Arrow

materialize to host language -> compute -> discard) is a general shape this post borrows without the paper's own unmeasured figures. - PAPER-4578 (V18, executable symbolic reasoning over Arrow +

Research Notes · Aug 27, 2026

Go and Look: Closing Your First Loop

Go and Look: Closing Your First Loop You wrote something down. A claim, a number that would settle it, and a date — committed somewhere you cannot quietly edit it.

Research Notes · Aug 26, 2026

Start Here: Your First Hypothesis, in One Sitting

60%, commit moved back 26 days, clean. Worse, the fix was already in the page and orphaned: Step 1 created a GitHub account, justified it as "where your timestamp comes

Research Notes · Aug 6, 2026

Adding a Relationship Type: Zero Lines of Rust — and the Two Costs That Aren't

Nobody needs another kanban board, and this post is not about one. We open-sourced arrow-kanban — an Arrow-native work-graph engine with Parquet persistence, typed relationships, and a NATS server mode

Research Notes · Jul 28, 2026

Graph Engineering Needs Graph Governance

The market spent July discovering that agents should think in graphs. It is right. The label is new — roughly three weeks old at meme scale — but the two