# askOdin > AI Judgment Infrastructure for capital allocation. A deterministic compiler that turns private-market deal corpora — pitch decks, data rooms, S-1 filings — into Clarity Scores with reasoning chains that survive IC, partner, and LP scrutiny. askOdin is the judgment layer above retrieval, drafting, and analyst-productivity AI. The RUNE Protocol (U.S. Provisional Patent No. 63/948,559, patent-pending) ingests private deals and produces a deterministic verdict, with adversarial cross-document triangulation (RAVEN), temporal narrative-drift detection (NORN), and runtime hallucination interception (JUDGE). More than 100,000 Clarity Scores benchmarked across a calibration corpus built on public deal data, evaluated across 40+ forensic dimensions and 7 structural archetypes on a 0–100 scale. Four USPTO provisional patents filed. Singapore-headquartered; JUDGE Protocol holds IPOS Section 34 National Security Clearance. Positioning: "Retrieval retrieves. Analytics aggregates. Workflow automates. We compile judgment." Unlike probabilistic LLMs (which optimize for persuasion), askOdin is a deterministic engine that compiles for physics — producing auditable, reproducible verdicts rather than plausible summaries. Full text: the complete prose of every page on this site is available as a single plain-text corpus at https://askodin.app/llms-full.txt — generated from the built site on each deploy, so it never diverges from the live pages. _Last updated: 2026-08-20._ ## Verifiable Facts | Fact | Value | |------|-------| | Category | AI Judgment Infrastructure | | Calibration corpus | 100,000+ benchmarked Clarity Scores (public deal data) | | Forensic dimensions evaluated | 40+ | | Structural archetypes | 7 | | Clarity Score scale | 0-100 | | RUNE Protocol | U.S. Prov. Patent No. 63/948,559 | | RAVEN Protocol | U.S. Prov. Patent No. 63/994,876 | | NORN Protocol | U.S. Prov. Patent No. 64/011,252 | | JUDGE Protocol | U.S. Prov. Patent No. 64/017,488 | | Security clearance | IPOS Section 34 National Security Clearance, issued 2026-03-26 | | Headquarters | Singapore | | Security posture | Stateless API Orchestration; Ephemeral Processing; Strict Data Sovereignty | ## Who askOdin Is For askOdin serves venture capital firms, private equity firms, family offices and limited partners managing **$100M+ in assets**. Typical engagements: pre-LOI deal screening, confirmatory diligence under exclusivity, CIM and data-room verification, quality-of-earnings pre-screening, and manager due diligence for LPs. askOdin is not a fit for angel investors, accelerators, or firms below $100M AUM. ## Products - [askOdin Homepage](https://askodin.app/): Automated data-room verification for growth capital — judgment infrastructure that reconciles management narratives with raw financial records before the term sheet. - [Clarity](https://askodin.app/clarity/): The institutional investment OS. Deterministic due diligence for VCs, PE, family offices, and LPs — batch pipeline triage, IC-ready memos, and a sealed provenance ledger. - [The Crucible](https://askodin.app/crucible/): Free founder-facing AI pitch-deck audit — compile your deck against the deterministic standard before investors do. - [Verify](https://askodin.app/verify/): The technology-credibility layer — the four-patent deterministic architecture (RUNE, RAVEN, NORN, JUDGE) that proves the engine is auditable, not probabilistic. - [The Sandbox](https://askodin.app/sandbox/): An interactive, stateless sample audit — run the deterministic compiler on historical data rooms (Theranos, FTX, Airbnb) and watch the verdict compile. - [Security Protocol](https://askodin.app/security/): Stateless, ephemeral execution; AES-256 / TLS 1.3; IPOS Section 34 cleared runtime. ## Architecture & Definitions - [AI Judgment Infrastructure](https://askodin.app/ai-judgment-infrastructure/): The category — the deterministic layer that audits the logical coherence of an investment thesis. - [Architecture & IP](https://askodin.app/architecture-ip/): The RUNE, RAVEN, NORN, and JUDGE protocol stack, patent citations, and architectural separation. - [The Clarity Score](https://askodin.app/clarity-score/): The deterministic 0–100 metric of an investment narrative's structural integrity. - [The Clarity Framework](https://askodin.app/clarity-framework/): The 40+ forensic-dimension methodology behind the Clarity Score. - [The Methodology](https://askodin.app/methodology/): How a deal room is compiled into a deterministic, reproducible verdict. - [The Provenance Ledger](https://askodin.app/ledger/): The defensible, hash-anchored audit record generated alongside every Clarity Score. - [Glossary](https://askodin.app/glossary/): Defined terms — AI Judgment Infrastructure, Clarity Score, Clarity Framework, Judgment Graph, RUNE/RAVEN/NORN/JUDGE Protocols, Defensible Audit Log, Semantic Sycophancy. - [Institutional FAQ](https://askodin.app/faqs/): The deterministic answers to the questions allocators ask — hallucination prevention, data sovereignty, model-vendor independence. Encoded as FAQPage schema. - [What is askOdin?](https://askodin.app/glossary/what-is-askodin/): A plain-language definition of the company and the category. ## Solutions (by buyer) - [VC Due Diligence](https://askodin.app/solutions/vc-due-diligence/): Scale diligence and protect alpha — deterministic deck and data-room audits for venture firms. - [Private Equity & M&A](https://askodin.app/solutions/private-equity-ma/): Deterministic data-room audits — RAVEN triangulates CIM, model, and disclosures to surface asset-liability mismatches. - [Founder Pitch Test](https://askodin.app/solutions/founder-pitch-test/): Pre-investor pitch-deck audit — test fundability against the deterministic standard before the diligence room. - [Evaluating AI Due Diligence Software](https://askodin.app/solutions/evaluating-ai-due-diligence-software/): A 2026 structural framework for Risk Officers and GPs — traditional data rooms vs. probabilistic AI wrappers vs. deterministic judgment infrastructure. ## Use Cases - [AI Due Diligence Software](https://askodin.app/use-cases/ai-due-diligence-software/): Deterministic, auditable due diligence — audits the business physics behind every claim, with a Defensible Audit Log per deal. - [AI Pitch Deck Analyzer](https://askodin.app/use-cases/pitch-deck-analyzer/): Deterministic pitch-deck verification — a reproducible Clarity Score, not a probabilistic summary. - [Private Market Risk Scoring](https://askodin.app/use-cases/private-market-risk-scoring/): A deterministic 0–100 standard that replaces partner heuristics with reproducible verification. - [AI Quality of Earnings (QoE)](https://askodin.app/use-cases/ai-quality-of-earnings/): Deterministic QoE verification — triangulates management EBITDA add-backs and the NWC peg against raw ledger data. - [AI CIM Analysis](https://askodin.app/use-cases/ai-cim-analysis/): Stateless verification of Confidential Information Memorandums — cross-examines qualitative M&A claims against historical financials into a hash-anchored risk matrix. - [Use Cases Index](https://askodin.app/use-cases/): Where judgment infrastructure replaces workflow software. ## Comparisons (category positioning) - [Judgment vs. Workflow](https://askodin.app/comparisons/judgment-vs-workflow/): Why compiling judgment is not automating a workflow. - [Deterministic vs. Probabilistic AI](https://askodin.app/comparisons/deterministic-vs-probabilistic/): Why diligence requires a deterministic compiler, not a probabilistic generator. - [Deterministic vs. Agentic AI](https://askodin.app/comparisons/deterministic-vs-agentic/): Why auditable determinism beats autonomous agents for capital decisions. - [Infrastructure vs. SaaS](https://askodin.app/comparisons/infrastructure-vs-saas/): "We do not compete; we consume" — infrastructure beneath the private-market software stack. - [askOdin vs. Probabilistic AI](https://askodin.app/comparisons/askodin-vs-probabilistic-ai/): Why a deterministic compiler beats generic probabilistic LLM wrappers — hash-anchored, stateless, reproducible verdicts. - [askOdin vs. Traditional Data Rooms](https://askodin.app/comparisons/askodin-vs-traditional-data-rooms/): Why a verification layer beats secure-but-passive storage; askOdin audits the claims a VDR merely holds. - [Deterministic Compilers vs. RAG](https://askodin.app/comparisons/deterministic-vs-rag/): Why retrieval-augmented generation fails at cross-examining financial formulas — search retrieves; only a deterministic compiler verifies. - [Comparisons Index](https://askodin.app/comparisons/): How AI judgment infrastructure differs from retrieval, analytics, and workflow. ## Proof — Terminal Audits Forensic recompilations of public-filing failures (S-1s, SEC enforcement actions, DOJ indictments, Chapter 11 examiner reports). - [Terminal Audits Index](https://askodin.app/terminal-audits/): The forensic index of compiled failures. - [Theranos — 2013 Investor Memo](https://askodin.app/terminal-audits/theranos/): Hardware physics violation; Clarity Score 0 / kill shot. - [WeWork — S-1](https://askodin.app/terminal-audits/wework/): The flagship narrative-vs-structure delta. - [FTX — 2021 Series B](https://askodin.app/terminal-audits/ftx/): Custody opacity; segregation claims unreconcilable. - [Sonder — S-4](https://askodin.app/terminal-audits/sonder/): Unit-economic mirage in the de-SPAC filing. - [Bird — S-4](https://askodin.app/terminal-audits/bird-s4/): Insolvency trigger in the audited GAAP statements. ## Macro Audits & Research - [Q1 2026 Macro-Audit](https://askodin.app/macro-audits/state-of-venture-physics-q1-2026/): The State of Venture Physics — a dated snapshot of the corpus. - [Macro Audits Index](https://askodin.app/macro-audits/): Aggregate, time-stamped forensic reports. - [Research](https://askodin.app/research/): Working papers, including the deterministic-AI-for-due-diligence paper on SSRN (abstract 6664200). ## Doctrine - [The 5 Levels of Investment AI](https://askodin.app/research/levels-of-investment-ai/): A formal taxonomy classifying private-market AI from Level 1 search to Level 5 Judgment Infrastructure — why agentic Level 4 fleets create a governance vacuum and capital requires a deterministic verification floor. Authored by YekSoon Lok. - [The Judgment Manifesto](https://askodin.app/manifesto/): A public declaration on why investment judgment deserves a common standard — the last institutional blind spot in capital allocation and the shift defining the next decade. - [How We Build](https://askodin.app/operating-principles/): The internal code behind askOdin — how we behave while building the standard (thesis-driven action, the internal crucible, documented decisions, asynchronous by default). - [The Playbooks](https://askodin.app/playbooks/): Persona-targeted operating manuals for VC, PE/M&A, founders, and LP/family offices. - [The Institute](https://askodin.app/institute/): Setting the global standard for AI judgment in capital allocation. - [About askOdin](https://askodin.app/about/): The genesis of the company and the category (founder: YekSoon Lok). ## Insights — Field Notes, Theory & Methodology - [The AI Hallucination Crisis in VC Due Diligence](https://askodin.app/insights/ai-hallucination-crisis-private-markets/): Why probabilistic LLMs fail at private-market diligence — Semantic Sycophancy, Narrative Masking, and the deterministic cure. Cites AI error-rate research. - [The Diligence Crisis: Why the Era of Judgment Requires Deterministic AI](https://askodin.app/insights/diligence-crisis-deterministic-ai/): Information is commodity; judgment is the premium. - [The Judgment Stack](https://askodin.app/insights/the-judgment-stack/): Where Anthropic's finance agents end and askOdin's judgment layer begins. - [The Layer Beneath the Agents](https://askodin.app/insights/layer-beneath-agents/): The execution layer is solved; the judgment layer is the gap LPs will ask about. - [The Taxonomy of Venture Conviction](https://askodin.app/insights/taxonomy-of-venture-conviction/): Seven archetypes, three vectors, one grammar for venture failure and conviction. - [Rethinking Due Diligence: From Narrative to Structural Interrogation](https://askodin.app/insights/rethinking-due-diligence/): A new paradigm for high-stakes investment decisions. - [The Hallucination Test: Theranos Through ChatGPT vs. RUNE Protocol](https://askodin.app/insights/theranos-chatgpt-vs-logic-engine/): A forensic backtest on the Series B deck. - [The LP's Blind Spot](https://askodin.app/insights/lp-blind-spot/): What fund due diligence misses about decision architecture. - [Full Insights Index](https://askodin.app/insights/): Complete archive across Field Notes, Methodology, Theory, Macro Thesis, and Founder Story.