askOdin Verify · The Public Record
Large language models optimize for persuasion.
askOdin compiles for physics.
Diligence does not reward a fluent narrative; it rewards execution you can repeat and check. Every askOdin verdict is computed outside the language model by a deterministic compiler, and anchored to the exact documents it read.
If you arrived here from a “Verified by askOdin” badge, this page is what that link attests — and how to check it without taking anyone's word for it, including ours.
FOUR U.S. PROVISIONAL PATENTS |63/948,559 · 63/994,876 · 64/011,252 · 64/017,488
THE PUBLIC RECORD
A claim you can check without trusting the claimant.
Founders who submit both a deck and a financial model, and whose numbers reconcile at 85 or above, earn a public Verify link. Some of them will put a “Verified by askOdin” badge on their own site that points back at it. If you have arrived here from one of those badges, this is what it means.
The link is anchored to SHA-256 fingerprints of the exact documents that were audited. You can confirm that the deck in your inbox is the deck that was scored, rather than a later revision wearing an older result. That is the whole point of a fingerprint: it fails loudly when the file changes.
What it attests is deliberately narrow — the claims in the deck reconciled against the numbers in the model, on the date of the audit, for those documents. It is not a verdict on the business, an endorsement, or a substitute for your own diligence. askOdin audits reasoning, not truth.
THE COMPILER ARCHITECTURE
Four layers between a claim and a verdict.
Standard AI summarises text. askOdin compiles assertions into a constraint-checked logic graph, and the verdict is computed outside the model. Four U.S. provisional filings cover the layers that do it.
RUNE Protocol™
Narrative compiler
Translates unstructured language into an executable, logic-validated dependency graph, anchoring every variable to its source text with a Brittleness Score.
RAVEN Protocol™
Cross-document triangulation
Checks a narrative assertion in a PDF against the embedded formula in a model, and preserves the contradiction rather than reconciling it away.
NORN Protocol™
Temporal drift detection
Reads chronologically sequential documents and isolates Narrative Inflation — rhetoric escalating while the underlying business physics degrade.
JUDGE Protocol™
Runtime circuit breaker
Intercepts a probabilistic hallucination before it reaches a verdict. When compiled logic violates a structural constraint, JUDGE halts and flags it rather than returning a confident number anyway.
// OBJECTION HANDLING
Technical FAQ
Why it isn't an LLM wrapper
Isn't this just a ChatGPT wrapper with a nicer UI?
No. A wrapper passes your prompt to a language model and formats the reply. askOdin restricts the language model to non-generative extraction, then compiles the extracted variables through a deterministic Go engine that enforces business physics. The model is the CPU; askOdin’s deterministic compiler is the operating system.
How is this different from RAG (retrieval-augmented generation)?
RAG retrieves text into a probabilistic model that still generates the answer — the verdict remains a generation. askOdin retrieves nothing into the judgment path: the RUNE Protocol compiles claims into a logic graph and a deterministic engine evaluates them. Retrieval retrieves; we compile judgment.
What exactly does the language model do versus the deterministic engine?
The language layer reads and extracts claims only — read-only and isolated. It never evaluates. A statically-typed Go engine performs every calculation and renders the verdict. The separation is the audit trail.
Determinism and proof
Can’t you just set temperature to 0 to make an LLM deterministic?
Temperature 0 only forces the model to emit its single most-probable token — it makes the output stable, not the reasoning mathematical, and model-version drift, tokenizer changes, and floating-point effects still move the result. More fundamentally, the verdict never touches the model: a deterministic Go engine evaluates the extracted claims outside the neural network. Reproducibility is a property of the architecture, not a sampling flag — identical inputs return an identical Clarity Score and an identical hash.
Is the output reproducible — same input, same score?
The compiled graph is. Once a document's variables are bound, the deterministic engine returns the identical Clarity Score every time it evaluates them — the logic path is invariant, not sampled. What we do not claim is that the extraction step is frozen forever: it runs on an external model, and model-version changes can move which variables come out of a document months later. That is why every audit is hash-anchored to the exact files and the Defensible Audit Log reconstructs any verdict down to the source cell or paragraph. You verify a past result by reading its record, not by hoping a re-run matches.
Are the patents granted or just provisional?
Four U.S. provisional patent applications are filed: RUNE (63/948,559), RAVEN (63/994,876), NORN (64/011,252), JUDGE (64/017,488). Stated plainly: filed and provisional, not granted.
Stop trusting probabilities. Deploy deterministic infrastructure.
Run the same compiler across your pipeline, before the committee votes rather than after the write-down.
Residency, retention and sub-processors, with each control labelled by what it actually is today — stated policy, implemented, or evidenced. No customer document enters any training corpus.