ByKaranteliPlatform Guide

Platform Scope, Workflow, and Usage Model

This public guide explains what ByKaranteli Terminal does, who it is for, and how the deterministic signal lifecycle is structured.

Back to Home Open Login

What the platform is about

ByKaranteli is a deterministic crypto futures signal terminal that runs 24/7 against Binance USDT-M perpetuals. It ingests candles, scores setups against a fixed rule set, produces entries with explicit take-profit and stop-loss targets, and evaluates the outcome of every signal candle-by-candle until one of four terminal states is reached: TP1, SL, TIMEOUT, or EXCLUDED. The engine never re-runs historical data to clean up its track record; every published number comes from a signal that was actually produced live.

Behind the scenes, a side-aware ML calibrator ensemble (trained separately for LONG and SHORT) prices each candidate signal's expected value in basis points, after fees, slippage, and funding. An autopilot controller nudges the quality and EV thresholds based on live statistics. If production drops below the target cadence, the thresholds relax; if the false-positive rate climbs, they tighten. Every threshold change is logged in the audit trail on the status page.

The platform's stated goal is transparency at a level competing signal services refuse to provide: published methodology, auditable decision logs, live-not-backtest evaluation, conservative tie-breaks, and an immutable public record of every incident that touched the signal stream. Whether that is valuable to you is a judgment only you can make, but you will not be making it in the dark.

Who should use it

ByKaranteli is designed for self-directed traders who already operate on Binance USDT-M perpetuals and want to layer a disciplined, systematic second opinion over their own decisions. The ideal user is comfortable reading PnL tables, understands concepts like profit factor and max drawdown, and can size positions using the Kelly-based risk guide. It is not a beginner's platform and it is not an automation. Every trade still needs to be placed manually in your broker account.

If you are new to crypto derivatives, read the risk guide and the methodology before signing up. Leveraged futures carry the risk of losing your entire position. Past performance, even when evaluated live and deterministically, does not guarantee future results.

Usage flow

Start on the public home page to see the live performance snapshot, the most recent closed signals, and the liquidation-pressure preview.
Browse /performance for aggregate net-PnL, win-rate, profit-factor and drawdown numbers across 30, 90, and 180-day windows.
Explore /symbols to rank the 500+ tracked USDT perpetuals by composite score, and click through to a per-symbol detail page for any ticker.
Read /methodology once. It explains the deterministic TP/SL evaluation, the post-fee cost model, and the ML calibrator ensemble in full detail.
Register an account to build a watchlist, receive a weekly digest of your tracked symbols, and link Telegram for threshold alerts.
Activate a subscription to unlock the terminal workspace: market cache, scanner feed, technicals, regime overlay, and the full liquidation map with 1H / 4H / 12H intervals.

Deterministic policy highlights

Eligibility contract: USDT quote + PERPETUAL contract type + TRADING status are all required before a symbol can enter evaluation. Enforced in code, not configurable at runtime.
Excluded means excluded: If a symbol fails the eligibility filter it is stamped EXCLUDED and never back-filled into outcome statistics. We do not let failed setups quietly disappear from the record.
Conservative tie-break: When both TP and SL trigger in the same candle, SL wins. Most competing platforms pick TP here to inflate win rate; we deliberately do the opposite.
Recompute is snapshot-based: Historical recomputes use immutable signal snapshots, not whatever the current rule set looks like. This prevents coverage drift from hiding past performance changes.
Symmetric sides: LONG and SHORT signals run through the exact same quality gate, EV threshold, and calibrator pipeline. Any perceived side bias comes from market conditions, not from an asymmetry we introduced.

Where to go next

The public surface has everything you need to evaluate the platform before committing to a subscription. /performance shows aggregate net PnL, win rate, profit factor, and max drawdown across 30 / 90 / 180-day windows. /symbols ranks every tracked USDT perpetual by composite score, and each ticker has its own detail page with a candle-level outcome history. /signals is the live closed-signal feed, newest first, filterable by LONG / SHORT. The methodology page explains the engine in full; the risk guide helps you size positions responsibly using live statistics.

When you are ready, create a free account and build a watchlist. The weekly digest and threshold alerts will give you a feel for how the engine behaves on your symbols of interest before you commit further.

Public FAQ

What makes ByKaranteli different from other signal providers?

Three concrete things. First, we publish our methodology in full: the eligibility contract, the tie-break rule when TP and SL trigger in the same candle (SL wins, conservative), and the exact fee, slippage, and funding model. Second, every number on the public site comes from signals that were produced live; we do not re-run historical data to manufacture clean statistics. Third, when we make a platform change (threshold tune, strategy rebalance), it is recorded in the incident log on /status, so the public record is auditable.

Who is this platform for?

Traders who want the research and discipline of a quant-style evaluation pipeline without building one themselves. Members typically already trade Binance USDT-M perpetuals and want a second opinion grounded in deterministic evaluation: clean entries, documented exits, post-cost PnL, and a track record they can verify. It is not a broker; you still execute trades in your own account at your own risk.

What is a 'deterministic' signal engine?

Deterministic means the same input always produces the same output, with no manual overrides. Every signal has one of four outcomes: TP1 (take-profit hit), SL (stop-loss hit), TIMEOUT (neither hit in the evaluation window), or EXCLUDED (the symbol failed the eligibility filter). Nothing is interpretive. When you read a win rate or a profit factor on ByKaranteli, it means the rule set ran end-to-end against the closed-candle sequence and produced that number. No human picked the good trades.

How do I know the numbers are not cherry-picked?

The evaluation engine runs 24/7 and writes outcomes the moment they are decided. The /signals page shows the most recent 24 hours of closed signals with their net-bps outcomes in the same moment those outcomes were stamped. The /performance numbers are an aggregation of that same live stream, not a re-run. You can also hit /api/public/performance and /api/public/recent as JSON endpoints if you prefer auditing the raw data yourself.

Does ByKaranteli evaluate every symbol on Binance?

Only USDT-M perpetuals whose Binance contract state is TRADING and whose symbol meets the eligibility filter. Stocks, FX, spot-only pairs, and expired / delisted contracts are excluded at the ingest stage and never reach evaluation. The filter is enforced in code; an operator cannot override it to produce optimistic statistics.

Can I access the raw data?

Yes. The /developers page documents four public JSON endpoints: aggregate performance, strategy leaderboard, recent closed signals, and per-symbol statistics. All are CORS-enabled with a 5-minute cache. No API key required for the public tier.

What happens to my watchlist when my subscription expires?

Your watchlist is preserved. You lose access to the terminal workspace and threshold alerts, but your saved symbols remain and re-activate automatically when you renew. Paywall reminder emails (7d, 3d, 1d buckets) tell you the exact expiry date so you can plan.