OQOmniQuantBeta₿ CRYPTO
Support

Frequently asked questions.

If your question isn’t here, email founder@omniquant.trade — we answer everything.

General

Is OmniQuant financial advice?+
No. OmniQuant is explicitly NOT financial advice, and we're legally clear about this. OmniQuant is an informational platform that provides: • Probability distributions and confidence intervals • Historical calibration metrics • Educational market analysis • Research-grade uncertainty quantification What we do NOT provide: • Buy/sell recommendations • Investment advice • Trading signals • Portfolio allocation guidance We are a decision support tool for sophisticated users who want to understand likelihoods and scenarios. We help you think probabilistically about markets—but the investment decisions are entirely yours. All users must acknowledge this in our Terms of Service before using the platform.
Is OmniQuant legal in my country?+
OmniQuant operates as an informational and educational platform, which allows us to serve users globally without requiring country-specific financial licenses. However, we recommend: • Users verify local regulations regarding market data consumption • Institutional users consult their compliance teams • Users in heavily regulated jurisdictions (e.g., certain EU countries) review their local laws regarding AI-powered financial tools We do NOT: • Execute trades on your behalf • Hold or manage your funds • Provide investment advice or recommendations This positions us as a research/intelligence tool rather than a regulated financial service. That said, regulations evolve—and we stay current with compliance requirements in our primary markets.
How is OmniQuant different from trading bots?+
We're fundamentally different in philosophy, purpose, and function: TRADING BOTS: • Execute trades automatically • Generate buy/sell signals • Often use leverage and margin • Goal: Maximize returns • Risk: Can lose your money OMNIQUANT: • Provides probabilistic intelligence • Explains scenarios and likelihoods • Never touches your funds • Goal: Improve your understanding • Risk: None (informational only) Trading bots try to beat the market. OmniQuant tries to help you understand the market. We don't tell you what to do. We help you understand what might happen and why. The decision to act (or not act) is entirely yours. Other key differences: • We explain our reasoning (Explainable AI) • We quantify our uncertainty (confidence intervals) • We track our calibration publicly (transparency) • We abstain when unsure (honesty over revenue) If you want a robot to trade for you, we're not that. If you want intelligence to inform your own decisions, that's what we provide.
How is OmniQuant different from other AI finance tools?+
Most AI finance tools fall into two categories: hype-driven "alpha generators" or black-box trading systems. We're neither. KEY DIFFERENTIATORS: 1. Calibration Is Measured, Not Claimed Well-calibrated would mean a 70% call resolves right about 70% of the time. We do not currently clear that bar — our published reliability curve says we are overconfident, and it says so on the public accuracy page rather than in a footnote. Treat the confidence number as a ranking, not as the literal odds. What we commit to is measuring the gap in the open and closing it. 2. Explainability We don't just give you a number. Our LLM-powered explanations tell you which signals contributed, what the model is "seeing," and why confidence changed. 3. Honest Uncertainty We show you confidence intervals, not point predictions. Markets are uncertain—our outputs reflect that reality. 4. Abstention Logic We're willing to say "we don't know" when our model shouldn't be confident. This is rare in AI tools that are incentivized to always provide an answer. 5. Post-Event Evaluation Every forecast is tracked against actual outcomes. We don't cherry-pick successes—we show you the full calibration picture. 6. No Trade Execution We're an intelligence layer, not a trading system. We inform your thinking; we don't make decisions for you.

Predictions

Why do you abstain from predictions sometimes?+
Abstention is a feature, not a bug. It's one of the most important signals we provide. We abstain when: • Model confidence is below acceptable thresholds • Market volatility exceeds training data distribution • Conflicting signals produce high uncertainty • Insufficient data exists for reliable forecasting • Regime change is detected Why this matters: • Most AI systems force an answer even when uncertainty is high • We believe "I don't know" is more valuable than a low-confidence guess • Abstention protects users from acting on unreliable forecasts To be precise about what abstention does and does not buy: it keeps us from publishing calls the model has no business making. It does NOT mean the calls we do publish carry a large measured edge — our high-confidence cohort runs only a few points above the published headline, and our own calibration work says stated confidence is a weak ordering rather than a reliable probability. The live numbers for both are on the accuracy page, and we would rather you read them than take a claim from us here. Think of it this way: A weather forecaster who says "I can't tell if it'll rain tomorrow" is more honest—and more useful—than one who guesses with false confidence.
What happens when the model is wrong?+
Models will be wrong. That's a mathematical certainty in probabilistic forecasting. Here's how we handle it: 1. Transparent Tracking: Every forecast is logged with timestamps and outcomes. We track accuracy, calibration, and confidence tier performance over time. 2. Calibration Adjustment: When we detect systematic bias — a confidence band resolving right less often than the number on it claims — we recalibrate. We publish the size of that gap on the accuracy page instead of describing it with an invented example, because a made-up pair of numbers in a FAQ tends to get quoted back as a real one. 3. Regime Detection: Our system detects when market conditions have shifted significantly from training data and signals reduced confidence. 4. Post-Mortem Analysis: Major misses are analyzed to understand what signals were missed or misweighted. 5. User Communication: We don't hide poor performance. Our calibration dashboards show exactly how well (or poorly) we've performed historically. The goal isn't to be right every time—it's to be honest about our uncertainty and improve over time.
How current are the predictions?+
Snapshots refresh continuously through the trading day. High-conviction calls are recomputed on a schedule, and your dashboard shows the freshness of every surface. Predictions are generated on trading days — markets closed means no new calls.

Accuracy

Can OmniQuant guarantee accuracy?+
No. And anyone who claims to guarantee market prediction accuracy is lying to you. Here's the honest truth about our system: What we CAN guarantee: • Transparency about our calibration metrics • Honest reporting of our historical performance • Continuous improvement based on outcome data • Clear communication when confidence is low What we CANNOT guarantee: • Future accuracy (markets are inherently uncertain) • Specific outcome correctness • Performance during black swan events • Accuracy on assets with limited training data Our models are DESIGNED to be well-calibrated — meaning a "65% probability" should resolve right about 65% of the time. Designed to be is not the same as measured to be, and today they are not: the published reliability curve shows stated confidence running ahead of the delivered hit-rate. We are telling you that here rather than letting you discover it on the chart. We publish the calibration curves and the error, unrounded, so you can evaluate the track record yourself instead of taking our description of it.
How do you measure accuracy?+

Directionally, on the real close-to-close move. A call is right if the price moves the way we said by its horizon, measured on official closes — not intraday ticks. Every prediction is locked before its outcome window opens, so it’s out-of-sample every time.

Two exclusions are worth knowing about, because they change the denominator. Moves too small to be a direction — smaller than the symbol’s own noise band — are thrown out of both the numerator and the denominator rather than being counted as wins; the band rule in force is: move smaller than the symbol's noise band, resolved at validation time: 0.5× the symbol's 20-day daily-return stdev clamped 0.1–5% where a volatility snapshot exists (about half of settled rows), otherwise a flat 0.3% default. The band is not scaled by horizon. That currently removes 18.1% of settled calls. And a confidence floor may exclude the calls the model itself scored lowest. No confidence floor is applied — every scored prediction is in that cohort, including the ones we were least sure of.

The published cohort definition, exactly as the engine reports it: Strict directional accuracy — a prediction is scored correct only when the symbol moved beyond its noise band and the model called that direction. Band-indeterminate outcomes (see indeterminate.band_description for the exact rule in force) are excluded from both the numerator and the denominator — not counted for or against. Cohort also excludes archived, abstained, out-of-universe, and shadow-challenger predictions. No confidence floor is applied: every scored prediction is in the cohort, including the ones the model was least sure of. Short (DOWN) calls are excluded: the platform publishes long calls only, so the track record describes the product actually sold. Short predictions made before 2026-08-23 remain in the database but are not scored here.

The full breakdown — per horizon, per market, direction split, and the misses — is on the track record.

Pricing

How do you make money long-term?+
Our business model is designed for sustainable growth, not hype-cycle exploitation. REVENUE STREAMS: 1. Subscription Revenue (Primary) • One plan — Pro, ₹499 / $9.99 a month, every market included • Free trial of every feature, no card required • Sustainable recurring revenue as we scale 2. Enterprise Licensing • Programmatic API access for institutional users • Custom calibration for specific asset classes • White-label intelligence for financial platforms 3. Data Products (Future) • Aggregated, anonymized forecast data • Calibration research datasets • Market regime analysis reports UNIT ECONOMICS: • Our infrastructure costs scale sublinearly • As the user base grows, cost-per-user decreases — letting us reinvest in research and accuracy rather than raise prices • We aim for profitability at 10,000+ subscribers
Why are you crowdfunding if the product already exists?+
Great question. Here's the honest answer: THE PRODUCT EXISTS, BUT GROWTH IS CAPITAL-INTENSIVE: What we've already built: • Core ML pipeline • Full-stack platform (Next.js + FastAPI) • Calibration and accuracy tracking • Data infrastructure across multiple markets • User authentication and subscription system What we need capital for: • Marketing & user acquisition • Additional engineering talent • Expanded data coverage (more assets, more frequencies) • Enhanced AI explanation features • Production-scale infrastructure scaling • Regulatory preparation for institutional clients WHY NOT VENTURE CAPITAL? • VCs often push for hyped messaging that conflicts with our philosophy • We want users—not investors—to be our primary stakeholders • Crowdfunding aligns incentives: supporters become users WHY NOW? • Product-market fit is validated • Technical foundation is solid • This is the growth phase, not the build phase WHAT BACKERS GET: • A fixed number of months of Pro — every market — plus a permanent supporter badge • A permanent supporter badge on the account • Not equity, not dividends, not voting rights, and no promised returns Crowdfunding isn't a sign of weakness—it's a strategic choice to grow with our community.Tiers and the full disclaimer are on the crowdfunding page.
What do I get on a paid plan?+
Daily high-conviction picks, deeper per-company intelligence (fundamentals, sentiment, news, governance and the “why” behind each call), watchlists with alerts — across every market, on one plan. See pricing.

Billing

Can I cancel anytime?+
Yes. Cancel from your account and access continues to the end of the paid period. Payments are handled by our payment partner; we don’t store your card details.
What is your refund policy?+
Payments on OmniQuant are final and non-refundable. That covers both of the things we sell: subscription plans, and one-time crowdfunding contributions (charged in INR via Razorpay). By completing a payment, you acknowledge and accept this policy. Please review your order carefully before confirming. This is intentional and product-driven: • Paid access is granted immediately on payment, so service consumption can't be retroactively verified • A crowdfunding contribution grants its full block of plan-months up front and never auto-renews — there is no unused remainder to return • Cryptocurrency payments are inherently irreversible on the blockchain • We may offer goodwill gestures (not refunds) for duplicate charges, fraudulent transactions, or extended platform outages You can cancel a subscription at any time to stop future renewals; access continues to the end of the billing period you already paid for. The complete terms and the exception process are shown at checkout before you confirm payment.
What happens if OmniQuant shuts down?+
We plan for continuity, but here's what happens in a worst-case scenario: IF OMNIQUANT SHUTS DOWN: 1. User Data Handling • All personal data would be securely deleted per GDPR/CCPA • Users would receive advance notice (minimum 30 days) • Export tools would be provided for any saved forecasts 2. No Financial Loss • We don't hold your money (we're not a custodian) • We don't execute trades on your behalf • Your investments are entirely separate from our platform 3. Subscription Refunds • Partial refunds for unused subscription time • Pro-rated based on remaining period OUR CONTINUITY PLANNING: • Infrastructure kept deliberately lean, so running costs stay low • Core infrastructure designed to be self-sustaining • Open-source fallback considered for core algorithms • Documentation sufficient for another team to operate OmniQuant is currently a solo-founder build. We are not going to quote you a runway figure we cannot show you: we're building for the long term, but no company is guaranteed to exist forever, and a small operation is less guaranteed than most. What we can commit to is the structure above — plus the refund terms in this answer — so that your exposure is limited regardless of our fate.

Markets

Which markets are covered?+
OmniQuant runs across seven market families, each with its own data feeds, models, and validation cohort: • US equities • India equities (NSE / BSE) • EU equities • APAC equities • Crypto • FX • Commodities One flat plan covers every market, and each market's calibration is tracked independently — a model that's well-calibrated on US large-caps may not generalise to APAC small-caps without separate validation.
Can institutions see different data than retail users?+
No. Same models, same data, same forecasts. OUR EQUITY PRINCIPLE: • Retail users see identical forecasts to enterprise clients • No early access for premium tiers • No hidden "alpha" reserved for institutions • Model updates roll out simultaneously to everyone WHAT DOES DIFFER BY TIER: • API rate limits (higher for enterprise) • Customer support SLAs (faster for enterprise) • Custom integration assistance (enterprise only) • Bulk export capabilities (enterprise only) • Historical data depth (varies by tier) WHAT NEVER DIFFERS: • Forecast outputs • Confidence assessments • Calibration metrics • Model version We believe the democratization of financial intelligence is essential. If we gave institutions better forecasts than retail users, we'd be perpetuating the information asymmetry that makes markets unfair. OmniQuant is built for a level playing field.

API & Enterprise

Are predictions logged and auditable?+
Yes, completely. Auditability is core to our credibility. WHAT WE LOG: • Every forecast generated (timestamp, asset, horizon, confidence) • Model version used • Input signals at prediction time • Actual outcomes when known • Calibration metrics computed WHY THIS MATTERS: 1. Prevents Cherry-Picking We can't hide bad predictions. Everything is timestamped and tracked. 2. Enables Calibration Analysis You can see our full performance history, not just highlights. 3. Regulatory Readiness If regulators ever require forecast logging, we're already compliant. 4. User Trust You can verify our claims about historical accuracy. ACCESS: • Users see their own forecast history in the dashboard • Aggregated calibration data is publicly visible • Enterprise clients can receive detailed audit reports • Third-party audits are welcomed and supported We believe radical transparency is our competitive advantage.
Do you offer API or institutional access?+
Yes — the top tiers include API access, and we support custom universes and desk deployments. Get in touch and we’ll scope it.

Security

How do you protect user data?+
Data security is foundational to our platform. Here's our approach: TECHNICAL SECURITY: • Encryption at rest (AES-256) and in transit (TLS 1.3) • No storage of financial credentials (we don't connect to brokerages) • MongoDB Atlas with enterprise security features • Least-privilege access controls, reviewed regularly • Dependency and code scanning in our build pipeline WHAT WE HAVE NOT DONE YET: • No third-party penetration test has been performed to date • No SOC 2 audit has been started; it remains an aspiration, not a roadmap item with a date We would rather tell you this than imply an assurance we have not obtained. OmniQuant is a solo-founder build; when either of the above changes, this answer changes with it. DATA MINIMIZATION: • We collect only what's necessary to provide the service • Email, hashed password, subscription status—that's it • No tracking of your external trading activity • No sharing with third-party advertisers GDPR & CCPA COMPLIANCE: • Right to access your data • Right to deletion • Right to data portability • Explicit consent for any data processing • No selling of personal information WHAT WE DO TRACK (transparently): • Your forecast views (to improve recommendations) • Model interaction patterns (to improve UX) • Feedback you provide (to improve calibration) You can request a full data export or deletion at any time.