Live Performance
Anonymised real-time tracking of three independent production accounts trading AlgoZilla’s Omniscius signals — current model Pythia v6.2. Updated every cron tick. Returns shown in percent only. Vertical markers in the chart show when each model version went live.
Cumulative Return — live
Anonymised — Portfolio A, B and C are independent production accounts trading AlgoZilla’s Omniscius signals — current model v6.2 (Pythia), with earlier trades under v5.0 Kairos, v4.2 Theros, v4.1 Katharsios, v4.0.1 Pythia and v4.0 Itzamná. Returns are computed in % since tracking start; absolute amounts and exact positioning are intentionally withheld. Last updated 12:10 UTC, Jun 20.
What you're looking at
Three real production accounts trading AlgoZilla’s Omniscius signals — current model v6.2 (Pythia), with earlier trades under v5.0 Kairos, v4.2 Theros, v4.1 Katharsios, v4.0.1 Pythia and v4.0 Itzamná. Every cron tick (5 minutes) the system refreshes the live balances and trade ledger, recomputes the metrics above, and stores them in the WordPress database. Real positions, real fees, real prices — not a paper-trade simulation.
To respect the traders we anonymise the accounts as Portfolio A, B and C. The Aggregate view combines both into a single book. We deliberately don't show absolute EUR amounts or exact entry prices for ongoing positions — only the percentage performance, win rate, drawdown, and trade timestamps. This protects the individual traders without compromising what makes the page useful: a transparent, real-time view of how the model performs in production.
Returns are computed against each account's initial capital — the EUR pot the bot was originally given to deploy. Max drawdown walks the full closed-trade equity curve plus current unrealised PnL, so it captures real peak-to-trough drops even before this dashboard existed.
Active Positions
| Portfolio | Coin | Entry | Held | Entry Price | Unrealized | Pulse Sentiment |
|---|---|---|---|---|---|---|
| A |
XLM
Stellar
|
12 May 2026 18:07 | 38d 18h | €0.138535 | +0.00% | CHARGING |
| A |
DOT
Polkadot
|
12 May 2026 17:07 | 38d 19h | €1.1212 | +0.00% | CHARGING |
| A |
ADA
Cardano
|
12 May 2026 15:07 | 38d 21h | €0.229915 | +0.00% | CHARGING |
| A |
PENDLE
PENDLE
|
11 May 2026 14:04 | 39d 22h | €1.6500 | -7.95% | AWAKENING |
| A |
FIL
FIL
|
11 May 2026 06:05 | 40d 6h | €0.976080 | -17.15% | CHARGING |
Recently Closed
| Portfolio | Coin | Closed | Held | Entry → Exit | P/L | Regime | Exit reason | Model |
|---|---|---|---|---|---|---|---|---|
| B |
DOT
Polkadot
|
17 May 2026 01:31 | 4d 8h | €1.1213 → €1.0839 | -3.55% | BEAR | trail | v5.0 KAIROS |
| A |
XVG
Verge
|
12 May 2026 18:12 | 3h 0m | €0.0030473 → €0.003058 | +0.33% | CRASH | signal | v5.0 KAIROS |
| C |
XRP
XRP
|
12 May 2026 18:02 | 3h 0m | €1.2151 → €1.2230 | +0.65% | BEAR | signal | v5.0 KAIROS |
| C |
BTC
Bitcoin
|
12 May 2026 18:02 | 19h 0m | €69,361.17 → €68,629.00 | -1.06% | BULL | signal | v5.0 KAIROS |
| B |
XRP
XRP
|
12 May 2026 17:06 | 2h 57m | €1.2263 → €1.2165 | -0.87% | BEAR | signal | v5.0 KAIROS |
| B |
BTC
Bitcoin
|
12 May 2026 17:04 | 18h 53m | €69,418.50 → €68,284.36 | -1.88% | BULL | trail | v5.0 KAIROS |
| A |
BTC
Bitcoin
|
12 May 2026 17:04 | 18h 53m | €69,417.50 → €68,294.00 | -1.85% | BULL | trail | v5.0 KAIROS |
| B |
ATOM
Cosmos
|
12 May 2026 08:10 | 57m | €1.7493 → €1.7401 | -0.77% | BULL | score_below_exit_threshold | v5.0 KAIROS |
| A |
WLD
Worldcoin
|
12 May 2026 04:08 | 59m | €0.236380 → €0.235310 | -0.47% | BEAR | score_below_exit_threshold | v5.0 KAIROS |
| A |
DOT
Polkadot
|
12 May 2026 02:10 | 12h 3m | €1.1577 → €1.1520 | -0.69% | BEAR | signal | v5.0 KAIROS |
| A |
INJ
Injective
|
12 May 2026 02:08 | 1d 13h | €3.5420 → €3.8801 | +9.12% | BULL | signal | v5.0 KAIROS |
| B |
DOT
Polkadot
|
12 May 2026 02:06 | 11h 56m | €1.1568 → €1.1520 | -0.42% | BEAR | signal | v5.0 KAIROS |
| A |
XVG
Verge
|
12 May 2026 02:05 | 1d 19h | €0.00317568 → €0.003127 | -2.14% | BEAR | trail | v5.0 KAIROS |
| B |
AAVE
Aave
|
12 May 2026 02:04 | 5h 57m | €86.9900 → €84.9400 | -2.30% | BEAR | signal | v5.0 KAIROS |
| A |
AAVE
Aave
|
12 May 2026 02:04 | 6h 1m | €86.5400 → €84.9300 | -1.88% | BEAR | signal | v5.0 KAIROS |
| B |
RNDR
Render
|
12 May 2026 02:03 | 7h 53m | €1.6858 → €1.6316 | -3.46% | BULL | trail | v5.0 KAIROS |
| A |
RNDR
Render
|
12 May 2026 02:03 | 7h 53m | €1.6853 → €1.6318 | -3.38% | BULL | trail | v5.0 KAIROS |
| A |
WLD
Worldcoin
|
11 May 2026 23:13 | 1d 15h | €0.228852 → €0.236260 | +2.80% | BEAR | score_below_exit_threshold | v5.0 KAIROS |
| C |
BTC
Bitcoin
|
11 May 2026 22:02 | 11h 59m | €68,811.10 → €69,495.97 | +1.00% | BULL | signal | v5.0 KAIROS |
| A |
FET
Fetch.ai
|
11 May 2026 20:08 | 4h 1m | €0.201540 → €0.201180 | -0.38% | BULL | signal | v5.0 KAIROS |
| B |
FET
Fetch.ai
|
11 May 2026 20:08 | 3h 57m | €0.201633 → €0.201180 | -0.23% | BULL | signal | v5.0 KAIROS |
| A |
FIL
FIL
|
11 May 2026 19:12 | 13h 6m | €0.976080 → €0.961400 | +0.00% | BULL | ladder_split_recovery | v5.0 KAIROS |
| B |
TAO
Bittensor
|
11 May 2026 19:09 | 3h 6m | €275.2100 → €276.1000 | -0.18% | BUBBLE | signal | v5.0 KAIROS |
| A |
TAO
Bittensor
|
11 May 2026 19:09 | 3h 6m | €275.1740 → €276.3300 | +0.02% | BUBBLE | signal | v5.0 KAIROS |
| A |
MOVR
Moonriver
|
11 May 2026 19:09 | 22h 7m | €2.1538 → €2.1981 | +0.63% | BEAR | trail | v5.0 KAIROS |
| B |
XRP
XRP
|
11 May 2026 19:08 | 3h 3m | €1.2558 → €1.2471 | -0.69% | BULL | exit_fp | v5.0 KAIROS |
| C |
XRP
XRP
|
11 May 2026 19:02 | 3h 0m | €1.2594 → €1.2482 | -0.89% | BULL | signal | v5.0 KAIROS |
| B |
LINK
Chainlink
|
11 May 2026 16:11 | 16h 5m | €9.1181 → €9.0005 | -1.29% | BULL | signal | v5.0 KAIROS |
| A |
LINK
Chainlink
|
11 May 2026 16:06 | 16h 1m | €9.1285 → €8.9626 | -1.82% | BULL | signal | v5.0 KAIROS |
| A |
APE
APE
|
11 May 2026 16:05 | 22h 0m | €0.135390 → €0.131300 | -3.61% | BULL | trail | v5.0 KAIROS |
Frequently Asked Questions
Are these real trades?
Yes — all three portfolios are real production accounts trading AlgoZilla’s Omniscius signals (current model v6.2 Pythia). Every fill, every fee, every drawdown shown here is from actual capital that the bot deployed. No paper-trade simulation, no cherry-picked period. The vertical markers in the chart above show when each model version went live, so you can see exactly which trades ran under which model.
Why anonymise the accounts?
Two reasons. First, privacy — the traders running these accounts didn't sign up for public exposure of their balances. Second, security: showing exact EUR positions on a live page invites unwanted attention. Percentages tell the same performance story without the risks.
How is return calculated?
Realised PnL (sum across closed trades) plus unrealised PnL (open positions marked-to-market against current market prices), divided by the account's initial capital — the EUR pot the bot was originally given to deploy. This is honest performance: no AUM-fluctuation noise, no inflated baseline.
How is Max Drawdown computed?
We walk every closed trade chronologically, building the cumulative-PnL curve, then append the current unrealised PnL as the “now” point. Max DD is the largest peak-to-trough drop on that curve, expressed as % of initial capital. This catches drawdowns that happened before this dashboard existed.
How often does the page update?
Every 5 minutes. A cron job refreshes the live balances and trade ledger, writes anonymised snapshots into the WordPress database, and the page auto-refreshes. The timer above the chart counts down to the next refresh in real time.
Why is the win rate < 50%?
A signal model can be highly profitable with sub-50% win rate if winners are larger than losers. Look at the avg-win vs avg-loss ratio on the per-coin signals pages. Recent live performance also reflects current market regime — the model retrains bi-weekly to adapt.
What does “v6.2 (Pythia)” mean?
Our current production signal model, live since 21 May 2026. Pythia v6.2 is the newest generation of the Omniscius machine-learning engine — three years of quantitative research distilled into one audited, reproducible code path. Where earlier versions leaned on hand-set thresholds, v6 lets the data decide: per-coin permutation tests determine which of 7,700+ candidate features per coin survive into the live model — statistically significant signals pass through, pure noise is filtered out automatically. Signal-aware stop logic respects the model’s conviction instead of firing blindly. Every live signal runs the exact same code path proven in backtest — leak-audited and validated through walk-forward testing across three Bitcoin halving cycles (2017–2026). Version history: v4.0 Itzamná (Apr 2026) confirmed the statistical edge; v4.0.1 Pythia, v4.1 Katharsios and v4.2 Theros refined exit logic and selective re-entry; v5.0 Kairos (8 May 2026) doubled the feature library and added per-coin 5-regime Pulse classification; v6.2 Pythia (21 May 2026) rebuilt feature selection around per-coin permutation testing. Read more →
What changed across versions?
v4.0.1 Pythia closed the backtest-vs-live exit-logic gap. Before Pythia the live executor had a simpler exit path: it acted on score-based signals only. The full backtest-grade exit logic — trailing stop on the high-water mark, hard stop on adverse moves, regime-driven exits, and a dedicated exit classifier that detects false peaks — only ran in shadow mode. Pythia promoted that whole stack into live. v4.1 Katharsios restored the calibrated hold-zone in sideways markets by removing a legacy macro-override that collapsed the entry-and-exit threshold gap to a single point — the model can now hold positions through normal sideways noise instead of churning. v4.2 Theros liberated the cooldown timer outside BULL, allowing selective re-entry. v5.0 Kairos extended that liberation to all five regimes (cooldown = 0 everywhere), fixed three structural bugs in the BUBBLE scorer that were silently misclassifying entries, expanded the feature library from ~200 to 506 features per coin per hour, added per-coin 5-regime Pulse classification (CRASH / BEAR / NEUTRAL / BULL / BUBBLE) on top of the existing macro Pulse, and switched the primary tuning KPI from a 4-bar entry-delay metric to a blended d0/d1 score that matches our actual ~30–90 second live execution latency. v6.2 Pythia (21 May 2026) rebuilt feature selection around per-coin permutation testing — each of 7,700+ candidate features must prove statistically significant predictive power before reaching the live classifier — and added signal-aware stop logic plus a full leak audit across three Bitcoin halving cycles. All versions share the same decision module across backtest and live, so “what backtest would have done” and “what the bot did” stay aligned.
Can I follow these signals myself?
Yes — the same Telegram signals that drive these portfolios are available to subscribers. Every alert includes entry price, stop-loss and take-profit. View pricing →
