Status
Current Project Status
This is not a server uptime dashboard. It is the current public state of the project — updated when the state changes, not on a fixed schedule.
Capability Maturity
What each part of the system actually is today, not what it is planned to become. Four states only — no partial credit.
| Live market data ingestion & observation | |
| Candidate generation (CG) | |
| Independent risk / lifecycle engine (RSKN) | |
| Virtual trade execution & lifecycle tracking | |
| Telegram delivery & closed-trade reporting | |
| Live public Reports (MAIN + SHADOW) | |
| ML evaluator architecture & validation | Active R&D |
| Multi-model evaluator architecture | Active R&D |
| Robustness & regime-conditioned validation methodology | Active R&D |
| Production-grade model promotion | Future |
| Controlled live-risk MVP | Future |
| TORYN token-gated access | Future |
| Real-money order execution | Disabled |
Current Focus
A bounded model- and method-selection research cycle for the ML evaluation layer has been completed: candidate approaches were compared under frozen, leakage-safe evaluation rules with independent audit. Completing this cycle does not authorize Production promotion — Production promotion remains a separate, explicitly unauthorized decision, and further bounded research cycles may follow. In parallel, the Telegram public-facing layer — including reporting on virtual trading history — continues to run and improve.
Current Limitations
- No model or method has been approved for Production trading. Everything observed today is virtual.
- A bounded model- and method-selection cycle has completed under frozen, leakage-safe rules with independent audit; the broader ML evaluation architecture remains active R&D and no model or method is treated as final.
- Live-market stability evidence beyond the current bounded verification is limited — long-run live-paper evidence is still being built.
- Asset and market coverage is intentionally narrow while the core pipeline is being proven.
- Public documentation intentionally omits proprietary thresholds, internal task identifiers and infrastructure detail.
Public Roadmap
Described at the level of product stages, not internal research task identifiers.
Active
- AI/ML research & validation
- Live-paper runtime
- Telegram public beta
- Public Build
Next
- Evaluator validation on fresh evidence
- Production-readiness validation
- Broader reliability and robustness work
Future
- Controlled live-risk MVP
- Expanded asset and reference coverage
- TORYN access layer