Skip to content

FEATURE_MATRIX.md

Latest commit

 

History

History
51 lines (43 loc) · 9.58 KB

FEATURE_MATRIX.md

File metadata and controls

51 lines (43 loc) · 9.58 KB
Raw

Feature Matrix

An honest status of what exists in this repository. Three states are used:

  • Implemented: wired end to end and exercised by tests or the running app.
  • Library: real code exists and can be imported and run, but it is not connected to the live API responses (which are seeded by default).
  • Infrastructure: configuration is present; standing it up is environment-specific.
  • Planned: not in the project.

Application

Capability Status Notes
Auth (register, login, refresh, profile) Implemented bcrypt + JWT; secret fail-closed in prod.
Portfolio (summary, positions, holdings, performance) Implemented Seeded data via service layer.
Strategies (list, detail, equity curve) Implemented Seeded metrics.
Risk (metrics, stress scenarios, radar) Implemented Seeded metrics.
Backtest (run, metrics) Implemented Real simulation: rl_agent/ml_alpha strategies train and run an ai_models agent/forecaster, momentum/mean_reversion run rule-based logic through BacktestEngine - all against a synthetic price series (see app/services/backtest_runner.py). Runs as a background task; poll GET /backtest/{id} for the result.
Market data (quotes, quote, history) Implemented Live via Yahoo Finance / Polygon, synthetic fallback.
Trading (orders, cancel) Implemented Simulated in-process; no live broker.
Research (papers) Implemented Static catalogue.
Alternative data (sources) Implemented Source registry with status and latency.
Web dashboard (all screens) Implemented React 19 + Vite + Tailwind.
Mobile app (all screens) Implemented Expo + React Navigation + Redux Toolkit.
Light / dark theme + toggle Implemented Persisted on device.
Offline demo fallback Implemented Both clients.
Contract / security / market-data tests Implemented In the backend suite.
Forecasting: transformer, attention, multi-horizon Implemented code/ai_models/forecasting, driven by ml_alpha backtests via app/services/backtest_runner.py.
RL agents: DDPG, PPO, replay buffer Implemented code/ai_models/agents, driven by rl_agent backtests via app/services/backtest_runner.py (PPO falls back to DDPG if stable-baselines3 isn't installed).
Trading / portfolio Gym environments Implemented code/ai_models/environments, used by rl_agent backtests for both training and evaluation.

Research library

Capability Status Location
Generative: GAN, regime model Library code/ai_models/generative
Bayesian Value at Risk Library code/backend/risk/bayesian_var.py
Stress testing, counterparty, aggregation Library code/backend/risk
Execution: liquidity forecasting, market impact, routing, OMS Library code/backend/execution
Analytics: alpha research, A/B testing, model validation Library code/backend/analytics
Data processing: caching, monitoring, parallel, pipeline, streaming Library code/backend/data_processing

Infrastructure

Capability Status Location
Dockerfiles + Compose (API, Postgres, Redis, Kafka) Infrastructure code/backend, infrastructure/docker
Kubernetes manifests (base, environments) Infrastructure infrastructure/kubernetes
Terraform modules (compute, database, monitoring, security, config) Infrastructure infrastructure/terraform
Ansible (inventory, playbooks, roles) Infrastructure infrastructure/ansible
GitHub Actions CI (4 jobs) Implemented .github/workflows/cicd.yml