Decoded Horizon: Strategy Simulator

Most leadership-assessment tools ask people how they think and trust the answer. Strategy Simulator watches what a person actually does inside a realistic high-stakes scenario — and scores the gap between the stated self-image and the observed behavior.

The problem it solves

The personality-assessment industry — DISC, MBTI, Big Five, Hogan, CliftonStrengths — is built almost entirely on self-report questionnaires. The buyer pays $50-500 per assessment for what is essentially a structured opinion the candidate has about themselves. A single executive mis-hire driven by that data costs $500K-2M in salary, severance, and rebuild time. A senior leader who tests as "strategic" on a self-report instrument and then freezes in a real crisis is a regular, expensive surprise across the organisations that buy these tools.

Existing business simulations (Capsim, Cesim, BTS) teach mechanics — P&L, market share, supply chains — but don't profile the decision-maker. Custom enterprise simulation programs from firms like BTS cost $50K-200K per engagement and still treat the simulation as a teaching event rather than a diagnostic instrument. There is no commercial tool that combines behavioral observation with strategic decision-making in the domains where the cost of bad strategic thinking is highest: DeFi, cybersecurity, financial regulation, adversarial operations.

Who needs this most

• HR Directors at 200-5,000-employee fintech, crypto, and cybersecurity firms who use DISC or Hogan today but cannot defend an executive-hiring decision when the candidate's first crisis exposes the gap between their assessment and their actual decision-making.
• Crypto and DeFi fund managers running 10-100-person teams who hire portfolio managers, governance analysts, and DeFi researchers and have no tool that understands their domain — every existing assessment was built for traditional finance.
• Cybersecurity firms and enterprise security-team leads whose red-team and blue-team capability assessments are conversation-driven and impossible to benchmark across hires or quarters.

The moment this hurts most: the week a senior strategic role needs to be filled, the candidate ranked highest on personality-test outputs has just stumbled in a panel-style scenario interview, and the hiring manager is being asked by the CEO to produce defensible evidence either way.

The solution — in plain terms

Strategy Simulator puts a person inside a realistic, high-stakes scenario — running a DeFi protocol through a governance attack, leading a crisis-response team during a breach, navigating an M&A defence, choosing how to respond to a regulator — and silently observes what they actually do. The platform sees what information they investigate first and what they ignore, how they allocate the resources they have, where they take risk and where they don't, and how they respond when a crisis hits mid-strategy. Out of that behavior comes a multi-dimensional profile that a self-report questionnaire cannot produce.

Operationally, every session runs through a 17-state machine: configure the scenario across six dimensions (domain, zone, complexity, role, time horizon, starting position), read the briefing, decide which hidden data to pay for (the Information Fog mechanic), write the strategy under time pressure, respond to an injected crisis if the complexity level warrants it, optionally face a three-round adversarial AI red-team, and receive a four-model score. Over five to ten sessions, the platform builds a Strategic Fingerprint — a stable longitudinal profile of how the person thinks, with eight axes that update via exponentially weighted moving average so recent behavior matters more.

For organisations, the same engine runs across a team, producing comparative profiles, surfacing collective blind spots, and feeding HTML reports defensible to internal stakeholders.

Value delivered — what you get

• Replace $500-2,000 personality assessments with evidence the candidate's strategic decision-making actually generates — every score is reproducible from the session log, not a recoded questionnaire answer.
• Quantify the gap between stated self-image and observed crisis behavior — the Shadow Index measures the distance between the moral zone the user chose and the moral zone their strategy actually occupied. That gap is the most powerful diagnostic signal of self-awareness level.
• Defend executive-hiring decisions to the CEO with audit-grade session logs — every decision, every resource allocation, every information request is timestamped and reconstructible.
• Assess domains existing tools refuse to touch — DeFi governance attacks, cybersecurity adversarial operations, grey- and black-zone strategic positioning, crypto-fund decision-making — all scenarios grounded in real mechanics.
• Build a multi-year Strategic Fingerprint per leader — track real growth in decision-making capability across 5-10+ sessions, not just self-perception updates between annual reviews.
• Run on infrastructure you control via the BYOK LLM gateway — bring Anthropic, OpenAI, Mistral, or run everything on local Ollama. No per-seat AI surcharge from the vendor and no strategy text leaving the customer's perimeter when on-prem is required.

Where it delivers outsized value

• Crypto and DeFi funds hiring portfolio managers, governance analysts, and DeFi researchers — the only assessment whose scenarios are grounded in actual DeFi mechanics rather than translated TradFi ones, so the output is defensible to investment committees.
• Cybersecurity firms and enterprise security teams — three-round adversarial AI red-teaming produces a resilience score that can be compared across red-team hires, training cohorts, or quarterly capability reviews.
• Fintech and regulated-finance organisations with on-premise data requirements — the Ollama integration keeps every strategy text, every resource decision, and every model call inside the customer's perimeter, which in-cloud assessment incumbents cannot match.

Distinctive features — why this over the alternatives

• Behavioral, not self-report — the user never tells the system what kind of strategist they are; the system infers it from resource allocation, information requests, and strategy text. The gap between the user's stated archetype (from the optional DAA self-report) and the observed archetype is itself a primary diagnostic.
• Four independent scoring models on every session — Strategic Score (0-100 quantitative across 9 components), Decision Architecture Profile (qualitative archetype), Shadow Index (moral-zone alignment), and Psycho-Behavioral Markers (resource-allocation pattern). One scenario produces four lenses, none of them substitutable.
• Information Fog as a diagnostic mechanic — what a person chooses to investigate (and what they ignore) reveals their information metabolism more honestly than a questionnaire about how they make decisions.
• Three-round adversarial AI mode — Probe, Escalation, Decisive — that attacks the specific vulnerability in the user's strategy. Produces a resilience score that cannot be gamed by pre-prepared answers.
• Strategic Fingerprint with EWMA — recent sessions weighted higher than old ones, so the longitudinal profile tracks real change instead of averaging it out. Status progresses through six stages from preliminary to mature.
• BYOK multi-provider LLM gateway — four active providers (Anthropic, OpenAI, Mistral, Ollama) with per-function routing, automatic fallback chains, and rule-based scoring always available as the last fallback. AI quality improves automatically as the underlying models improve.

Under the hood — built to last

The backend is a Fastify modular monolith on Node 20 and TypeScript, with Prisma against PostgreSQL 16 — well-supported foundations chosen so the platform still runs cleanly in five years. The 14 internal modules each map 1:1 to a future microservice and communicate by service imports rather than HTTP, so extraction is straightforward when scale demands it but the development loop stays simple today. The frontend is React 18 with Vite, Tailwind, Zustand, and Recharts. Everything ships in Docker; the full stack stands up on a single VPS with ./setup.sh, runs offline on customer infrastructure when needed, and has no SaaS dependency on the critical path. JWT RS256 auth, bcrypt(12) password hashing, AES-256-GCM for stored LLM keys, and a soft-deletion audit trail from day one.

Current maturity

All six implementation phases — Foundation, Core Engine, Intelligence Layer, Assessment, Enterprise, and Polish — are code-complete. The codebase is 173 TypeScript source files totaling roughly 42,700 lines across the four monorepo packages (shared, database, api, web). The database has 13 Prisma models with three migrations applied and 18 seeded scenarios (6 Arena tier + 12 Command tier). The dev environment is verified end-to-end: API healthy, Postgres seeded, frontend served, architect credentials working. The build was done as a focused sprint starting 2026-02-26, with milestone commits rather than micro-commits; the git history is short by design. Remaining work before paid launch: end-to-end integration testing with real LLM API keys, a security audit, production Docker with nginx and SSL, and a CI/CD pipeline. The platform is feature-complete BETA — usable end-to-end, not yet billed as a paid product.

Roadmap — what's next

The immediate revenue path runs through three milestones. First, paid DAA launch as a standalone $49 assessment that doubles as the marketing funnel for the simulation tiers — modeled on how DISC's free archetype quiz became the global lead-generation engine for the paid assessment market. Second, Command tier ($49/month subscription) for individual strategists, DeFi fund managers, and cybersec professionals — the first recurring-revenue product. Third, Sovereign tier ($99-149/seat/month, 10+ seat minimum) targeting fintech firms, crypto funds, and cybersec organisations with team analytics, custom scenarios, on-premise Ollama deployment, and HTML report exports.

The longer arc is a behavioral data moat. Every completed session adds to a calibration corpus that no competitor can match without comparable session volume — and the Strategic Fingerprint becomes more valuable per user the longer they stay on the platform, which compounds retention. Methodology licensing to executive coaches, HR consulting firms, and HR-tech platforms opens after the core paid product proves out, with a clear acquisition path to assessment incumbents (SHL, Hogan, Korn Ferry) once the methodology is validated across thousands of sessions.

Working with the architect

Strategy Simulator is available for three engagement modes. An HR organisation, crypto fund, or cybersec firm with a hard need the public product cannot yet serve can commission a custom build — domain-specific scenario packs, custom scoring rubrics, integration with an existing hiring or HR stack. A consulting firm or executive coach can engage in strategic advisory on assessment methodology, behavioral-scoring rubric design, or how to operationalize a Decision Architecture framework inside their own practice. And enterprise customers with multi-team needs can engage as a first Sovereign-tier tenant with on-premise deployment, custom scenarios, and integrated DAA + simulation reporting. Reach out via sintegrium.io or LinkedIn for a 30-minute scoping call.

---

Built by Yurii Staryk · Solution Ecosystem Architect