Reference cases · TickSense's track record
A small set of signature events that demonstrate what TickSense does — both catch (eruptions correctly identified within seconds) and avoid (fakes correctly flagged as incoherent). All timestamps are verifiable.
2026-05-04
RLYB — caught the eruption within 1 second
RLYB gapped from $9.41 → $12.88 in 3 minutes. The volcano fired an eruption event at 04:05:47 ET as the move started. By the time a human glanced at the chart, the system had already entered and exited.
Evidence (verifiable timestamps)
- 08:01 UTC $10.56 (gap up appears)
- 08:02 UTC $12.00 (+30% in 60 seconds)
- 08:04 UTC $12.88 high ← peak
- 08:05:47 UTC ERUPTION fired rvol 3.6x OBV BULLISH
- 08:06 UTC human notices
The volcano caught the move the moment its mechanics confirmed. Anyone watching their chart manually would have arrived 30-60 seconds late.
2026-05-03
PN — correctly avoided despite +294% headline
PN appeared to print a monster pump (+294% visible on the chart). Visually identical to a real eruption. The volcano refused to fire — flagged as incoherent due to bearish OBV during the price spike. Within 6 minutes the move collapsed -10%.
Evidence (verifiable timestamps)
- PN price: $1.42 → $6.01 (+294%)
- OBV: BEARISH throughout the spike (cumulative selling > buying)
- Volcano event: stale=True reason: incoherent (mechanics don't match price)
- Within 6 minutes: faded to $5.42 (-10% off peak)
- Within 30 minutes: dropped from active monitoring entirely
The volcano's discrimination layer turned 'looks like a trade' into 'stand down.' This is the half of the signal you won't get anywhere else.
2026-04-23
AKAN — full-stack proof of concept
First clean end-to-end trade. The volcano fired an eruption event, downstream bot entered, held for 10 minutes, exited at +27.72%. Operator's parallel manual trade got +1.39% on the same ticker.
Evidence (verifiable timestamps)
- AKAN eruption event timestamped, verifiable
- Bot exit at +27.72% via 10-minute hold
- Manual entry on same ticker: +1.39% (held shorter, exited earlier)
- 20× difference between automated and discretionary execution on identical signal
The volcano's signal is structured enough for autonomous execution. Same signal, mechanical execution captures 20× the discretionary outcome.
2026-04-27
HTCO — three independent bots, one signal
HTCO triggered. Three independent bot architectures (Tenacity A, Tenacity B, Shadow) — running different exit logic, different entry confirmations — all caught the same trade simultaneously.
Evidence (verifiable timestamps)
- Three different bots, three different strategies
- All entered HTCO within seconds of each other
- All caught approximately the same +25% move
- Confirms the underlying signal is real, not bot-specific quirk
When three independent gates converge on the same trade, the signal is doing the work — not the bot.