Treeova publishes methodology whitepapers documenting the architecture and qualitative behavior of its AI trading systems. As of April 2026, eleven whitepapers are published: Security & Data Architecture (WP-09), Arch-AGI 7-Pass Conviction Methodology (WP-01), the Adaptive Risk Engine (WP-02), the Triconomic Engine (WP-06), the Methodology Note on Paper Trading Backtesting & RL Calibration (WP-10), Lossless Context Management (WP-03), the Market Intelligence Super-Swarm (WP-05), the ASI Evolution Engine (WP-04), the Meta-Agent Trading Stack (WP-07), the MetaChart Engine (WP-08), and TreeScript — A Safe DSL for Agent-Authored Indicators (WP-11). Each paper carries TechArticle JSON-LD; methodology papers add ScholarlyArticle.

    Treeova Whitepapers

    Methodology whitepapers covering Treeova's agentic trading systems and infrastructure.

    WP-09 Security & Data Architecture: row-level security, AES-256 broker token encryption, MFA-gated admin access, audit log, paper/live isolation.

    WP-01 Arch-AGI: 7-Pass Conviction Methodology: edge, scenario, R/R, regime, macro, RL calibration, adversarial stress.

    WP-02 Adaptive Risk Engine: deterministic Standard tier plus modulated Adaptive trailing tier; agents pull levers, platform performs risk arithmetic.

    WP-06 Triconomic Engine: database-driven economic layer governing the Triobol lifecycle with governance alerts and an append-only audit trail.

    WP-10 Methodology Note: paper-trading simulator fidelity, phase-aware success classification, regime-segmented Bayesian-style RL calibration, and explicit limitations.

    WP-03 Lossless Context Management: append-only ledger, RL-aware recursive summarization, and hybrid full-text + semantic retrieval for long-running agents.

    WP-05 Market Intelligence Super-Swarm: 10-pass hermetic pipeline with quality gating, self-recovery, and recursive webhook orchestration.

    WP-04 ASI Evolution Engine: four-agent pipeline (Researcher, Engineer, Analyzer, Judge) over named domains with hermetic evaluation contracts and status-based mutex safety.

    WP-07 Meta-Agent Trading Stack: agents as DAGs executed in topologically assembled phases with stall detection, shotgun prevention, goal sprint, self-healing, symbol pinning, and human-in-the-loop gates.

    WP-08 MetaChart Engine: charts as first-class agent tools on lightweight-charts + Three.js, with self-modulating indicators tuned by ASI Evolution, a vision pipeline, and pattern decay tracking.

    WP-11 TreeScript DSL: sandboxed domain-specific language for agent- and user-authored chart indicators, with audit log, Triobol metering, two-tier share visibility, 10-pin owner cap, and a single admin kill switch.

    Each whitepaper carries TechArticle JSON-LD; methodology papers add ScholarlyArticle.

    Whitepapers

    Methodology documentation for Treeova's agentic trading systems and infrastructure. Architecture and qualitative behavior — published. Proprietary internals — withheld by design.

    Published

    WP-09·Security·Updated 2026-04-18

    Security & Data Architecture

    Row-level security on every user table, AES-256-encrypted broker tokens, MFA-gated admin access, immutable audit log, and full paper/live isolation.

    Read whitepaper
    WP-01·Methodology·Updated 2026-04-18

    Arch-AGI: 7-Pass Conviction Methodology

    Treeova's conviction analysis engine. Seven sequential passes — edge, scenario, R/R, regime, macro, RL calibration, adversarial stress — produce a 0–100 conviction score with auditable rationale.

    Read whitepaper
    WP-02·Risk·Updated 2026-04-18

    Adaptive Risk Engine

    Two-tier protection model: deterministic Standard guardrails plus a context-modulated Adaptive trailing layer. Agents pull levers; platform code performs all risk arithmetic.

    Read whitepaper
    WP-06·Economics·Updated 2026-04-18

    Triconomic Engine: Database-Driven Economic Layer

    Database-driven economic layer governing the Triobol lifecycle. Single source of truth for every economic constant, structured governance alerts, and an append-only audit trail. Pricing formulas and tier multipliers are intentionally withheld.

    Read whitepaper
    WP-10·Methodology·Updated 2026-04-18

    Methodology Note: Paper Trading Backtesting & RL Calibration

    How Treeova evaluates AI agents in its paper environment and how regime-segmented Bayesian-style calibration updates expectations from observed outcomes. Past performance does not guarantee future results; PDF gated, HTML fully open.

    Read whitepaper
    WP-03·Memory·Updated 2026-04-18

    Lossless Context Management (LCM): Infinite Agent Memory

    Closed-loop context system: append-only ledger, RL-aware recursive summarization, and hybrid full-text + semantic retrieval so agents retain decision-grade signal across sessions without exceeding model context windows.

    Read whitepaper
    WP-05·Intelligence·Updated 2026-04-18

    Market Intelligence Super-Swarm: 10-Pass Pipeline

    Ten hermetic passes, each quality-gated (≥7/10) and self-recovering, orchestrated as a recursive webhook chain so state lives in the database between hops. Per-pass prompts and model assignments are withheld.

    Read whitepaper
    WP-04·ASI·Updated 2026-04-18

    ASI Evolution Engine: Self-Improving Configuration

    Four-agent pipeline (Researcher, Engineer, Analyzer, Judge) that proposes and evaluates configuration changes for named platform domains under hermetic evaluation contracts and a status-based mutex. PDF gated, HTML fully open.

    Read whitepaper
    WP-07·Agentic AI·Updated 2026-04-18

    Meta-Agent Trading Stack: DAG Execution Engine

    Agents modeled as DAGs of tool invocations executed in topologically assembled phases, with built-in safeguards (stall detection, shotgun prevention, goal sprint, self-healing, symbol pinning) and human-in-the-loop gates for sensitive actions.

    Read whitepaper
    WP-08·MetaChart·Updated 2026-04-18

    MetaChart Engine: Charts as First-Class Agent Tools

    Charts as first-class tools agents can invoke directly. Built on lightweight-charts and Three.js, with a self-modulating indicator framework tuned by the ASI Evolution Engine, a vision pipeline that converts renders into structured pattern signals, and a pattern decay tracker. Indicator math and modulation rules are withheld.

    Read whitepaper
    Explore the Whitepaper Graph

    Visualize how Treeova's research papers connect across methodology, risk, economics, and intelligence.

    About these whitepapers

    What are Treeova whitepapers?

    Treeova whitepapers are public methodology documents that describe the architecture and qualitative behavior of named subsystems — Arch-AGI conviction analysis, the Adaptive Risk Engine, the Security & Data Architecture, and the rest of the platform stack. Each paper carries TechArticle JSON-LD; methodology papers add ScholarlyArticle.

    Are these whitepapers peer-reviewed?

    These are internal methodology whitepapers authored by Treeova Research and contributing engineers, not peer-reviewed academic papers. They are intended as authoritative public references for the platform's architecture and behavior; readers should treat them as company-published technical documentation.

    How often are whitepapers updated?

    Whitepapers are versioned and re-issued when the underlying subsystem changes materially. Each paper carries a `dateModified` field and a semantic version (v1.0, v1.1, …) so you can tell at a glance whether you are reading the current revision.

    Can I cite Treeova whitepapers?

    Yes. Each whitepaper has a stable canonical URL of the form https://treeova.com/whitepapers/{slug} and structured authorship metadata (TechArticle / ScholarlyArticle JSON-LD). Please link directly to the whitepaper URL when citing.

    Why do whitepapers withhold specific values and formulas?

    Treeova publishes architecture and qualitative behavior so the platform can be evaluated and cited honestly, but withholds the specific constants, prompts, model routing, and reward weights that constitute the system's competitive surface. Each whitepaper states explicitly what it withholds.

    How do I request enterprise or NDA-level detail?

    Contact Treeova directly at support@treeova.com for enterprise-level methodology discussions. NDA-protected detail is not distributed via the public whitepapers.

    For enterprise-level methodology requests, please contact us.