foxhunt/UNUSED_FEATURES_QUICK_REFERENCE.md

UNUSED FEATURES - QUICK REFERENCE

Last Updated: 2025-10-16 Overall Production Readiness: 95% System Status: ✅ READY FOR PAPER TRADING

---

🎯 TOP 5 FEATURES NEEDING INTEGRATION

1. Streaming Tuning Progress (tune_stream)

• Status: ✅ Implemented, ❌ Disabled
• Why: Waiting for API Gateway streaming support
• Fix Time: 2-3 hours
• Location: tli/src/commands/tune_stream.rs
• Impact: Users get better UX for long-running tuning jobs
• Command to Enable:

  # In tli/src/commands/mod.rs, uncomment:
  pub mod tune_stream;
  

2. TGNN (Graph Neural Networks)

• Status: ✅ Implemented (40%), ❌ Not Integrated
• Why: No order book data (Level 2 not in DBN files)
• Fix Time: 4-6 weeks (data + integration)
• Location: ml/src/tgnn/
• Impact: Advanced market microstructure insights
• Blocker: Need tick-by-tick order book snapshots (external data)

3. S3 Checkpoint Storage

• Status: ✅ Implemented, ⚠️ Manual activation
• Why: Not needed for local training, AWS account required
• Fix Time: 1-2 hours
• Location: storage/Cargo.toml feature flag
• Impact: Safe long-running job checkpoints
• Enable:

  storage = { path = "storage", features = ["s3"] }
  

4. ArrayFire GPU Library

• Status: ❌ Unused
• Why: Candle-core used instead
• Fix Time: 5 minutes (DELETE)
• Location: ml/Cargo.toml line 95
• Impact: Removes unused dependency warning

5. Advanced Labeling Strategies

• Status: ✅ Implemented (90%), ⚠️ Partially used
• Why: Already built, available if needed
• Fix Time: 3-4 hours analysis
• Location: ml/src/labeling/
• Features: triple-barrier, meta-labeling, GPU acceleration

---

📊 QUICK STATUS TABLE

Feature	Status	Priority	Effort	Blocker
tune_stream	Implemented	MEDIUM	2-3h	API Gateway
TGNN	Partial	MEDIUM-HIGH	4-6w	Order Book Data
S3 Storage	Implemented	MEDIUM	1-2h	AWS Account
ArrayFire	None	LOW	5m	N/A
Liquid NN	Complete	HIGH	0h	N/A
Labels	Implemented	LOW	3-4h	None
Regime Detection	Partial	LOW	2-3h	None

---

⚠️ CRITICAL FINDINGS

1. petgraph not feature-gated - Always compiled for TGNN even if unused
2. tune_stream commented out - Clear blocker noted in code
3. ArrayFire unused - Should be removed
4. Duplicate dependency versions - axum (0.7.9 vs 0.8.6), minor risk

---

✅ WHAT'S PRODUCTION READY

• ✅ Core trading engine
• ✅ ML models (MAMBA-2, DQN, PPO, TFT)
• ✅ Risk management
• ✅ API Gateway (22/22 gRPC methods)
• ✅ Monitoring (Prometheus/Grafana)
• ✅ E2E tests (22/22 passing)
• ✅ Paper trading validation
• ✅ GPU training (RTX 3050 Ti CUDA)

---

🚫 BLOCKERS FOR 100%

1. Order Book Data - TGNN needs Level 2 snapshots (not in DBN)
2. API Streaming - tune_stream needs gRPC server support
3. AWS Provisioning - S3 needs credentials

---

💡 NEXT STEPS

This Sprint (2-3 days):

1. Remove ArrayFire
2. Feature-gate petgraph
3. Document features

Next Sprint (1-2 weeks):

1. Check if API Gateway supports gRPC streaming (for tune_stream)
2. Provision AWS account (for S3 storage)

Future (4-6 weeks):

1. Acquire Level 2 order book data
2. Integrate TGNN training pipeline

---

FILE LOCATIONS

Core Implementation Files:
├── tli/src/commands/tune_stream.rs          (264 lines, ready to enable)
├── ml/src/tgnn/                              (6 modules, needs data)
├── storage/Cargo.toml                        (S3 feature flag)
├── ml/Cargo.toml                             (ArrayFire line 95, petgraph)
└── ml/src/labeling/                          (Advanced label strategies)

Documentation:
├── COMPREHENSIVE_UNUSED_FEATURES_AUDIT.md    (Full report)
└── UNUSED_FEATURES_QUICK_REFERENCE.md        (This file)

---

🎓 EXAMPLES

Enable tune_stream

// In tli/src/commands/mod.rs
pub mod tune_stream;  // Uncomment this line

// In tli/src/commands/tune.rs, add command:
Command::Watch(job_id) => {
    tune_stream::watch_tuning_progress_streaming(
        &config.api_gateway_url,
        &jwt_token,
        &job_id
    ).await
}

Enable S3 Storage

# In storage/Cargo.toml
[features]
default = ["s3"]  # Already set

# In services/ml_training_service/Cargo.toml
storage = { path = "../../storage", features = ["s3"] }

Remove ArrayFire

# In ml/Cargo.toml
# Delete this line:
# arrayfire = { version = "3.8", optional = true }

Feature-gate TGNN

# Add to ml/Cargo.toml [features]
[features]
graph-models = ["petgraph"]

# In ml/src/lib.rs
#[cfg(feature = "graph-models")]
pub mod tgnn;

---

📞 QUICK LINKS

• Full Audit: /home/jgrusewski/Work/foxhunt/COMPREHENSIVE_UNUSED_FEATURES_AUDIT.md
• ML Crate: /home/jgrusewski/Work/foxhunt/ml/Cargo.toml
• Storage: /home/jgrusewski/Work/foxhunt/storage/Cargo.toml
• TLI: /home/jgrusewski/Work/foxhunt/tli/src/commands/mod.rs
• CLAUDE.md: Project status and configuration