Introduction
DepthSight is an open-source platform for algorithmic cryptocurrency trading that offers a visual, no-code approach to building automated strategies. The platform provides a drag-and-drop strategy builder, AI-powered assistance, and enterprise-grade infrastructure designed for both individual traders and multi-user SaaS deployments. Unlike traditional trading bot frameworks that require Python coding, DepthSight enables users to design, backtest, and deploy trading strategies through a graphical interface.
Key Features
- Visual Strategy Builder: Drag-and-drop node graph interface with 40+ logic blocks for creating complex trading strategies without writing code
- AI Co-Pilot: Generate strategies from text prompts or chart screenshots; analyze live trades and backtests with AI assistance
- Dual Backtesting Engine: Fast vector engine for prototyping combined with candle/tick-level simulation for precise execution testing
- Dynamic Risk Management: Adaptive position sizing and risk parameters that respond to real-time performance per trading pair
- Multi-Exchange Support: Native integration with Binance and Bybit, with Bitget, OKX, Gate.io, and BingX in active development
- Market Scanner: Real-time market scanning tool for identifying order flow anomalies and receiving Telegram notifications
- Multi-Tenant Architecture: Isolated execution environments with per-user quotas and plan-based permissions
- Mobile PWA: Progressive Web App for monitoring positions and receiving alerts on mobile devices
- Genetic Optimization: Evolve strategy parameters through genetic algorithms with dynamic gene pools
- ML Pipeline: Online learning (River) and batch models (scikit-learn, XGBoost, LightGBM) for signal confirmation
Use Cases
DepthSight suits traders who want to automate cryptocurrency strategies without deep programming knowledge. The visual builder appeals to discretionary traders looking to codify their manual strategies into systematic rules. Active scalpers benefit from the real-time market scanner that monitors order book anomalies and sends Telegram alerts when opportunities arise.
For developers and SaaS entrepreneurs, the platform provides a complete "SaaS-in-a-box" architecture that can support hundreds of users with isolated execution environments. This makes it viable for launching commercial trading platforms similar to 3Commas or Veles.
Quantitative traders appreciate the dual backtesting engines-quick vectorized tests for idea validation followed by detailed candle-level simulations for accuracy. The genetic optimization and machine learning pipelines support systematic strategy development and refinement over time.
Pricing
DepthSight offers a self-hosted deployment option that runs 100% free on your own VPS infrastructure. Users can deploy the entire platform stack using a single shell script on Ubuntu 22.04+ servers with a minimum of 6 CPU cores and 16 GB RAM.
The project also provides a managed cloud deployment option for users who prefer not to manage their own infrastructure. Specific pricing tiers and feature comparisons for the managed offering are not publicly detailed on the website-interested users should inquire directly.
A Bybit exclusive promotion offers 5 free bots for real trading, though the specific terms and limitations of this offer require verification with the platform.
User Experience and Support
The platform emphasizes accessibility through its visual strategy editor. Users describe trading ideas in plain language, and the AI assistant assembles complete multi-stage strategies from these inputs. The interface uses React with shadcn/ui components, providing a modern, responsive experience.
Documentation appears comprehensive, covering quick start guides, architecture deep-dives, trading engine specifics, and API references. The interactive API documentation is available locally after deployment. Support channels include documentation, Telegram notifications for the scanner tool, and presumably community support given the open-source nature of the project.
The platform architecture decouples the API from the trading engine using Redis command channels, meaning API restarts do not disrupt active trades-a design consideration that benefits uptime and reliability.
Technical Details
DepthSight runs on a distributed architecture of 8 containerized services orchestrated by Docker Compose:
- Backend: FastAPI + Uvicorn with async-first design and automatic OpenAPI documentation
- Trading Engine: Python 3.11+ asyncio for concurrent exchange operations
- Database: PostgreSQL 15 with SQLAlchemy and Alembic migrations
- Caching/Messaging: Dual Redis 7.2 instances-one for persistent state (sessions, rate limits, Celery) and a separate volatile instance for high-frequency market data fan-out
- Task Queue: Celery 5.5 with prefork for CPU-bound backtests and optimization
- ML/AI: River for online learning, scikit-learn/XGBoost/LightGBM for batch inference, Google Gemini for the AI co-pilot
- Frontend: React 19 + Vite + Tailwind CSS with dnd-kit for drag-and-drop functionality
- Exchange Integration: CCXT 4.4 for unified API access across supported exchanges
- Optimization: DEAP for genetic algorithms, Optuna for Bayesian optimization, Numba for JIT compilation
The fan-out consumption pattern for market data means 100 users trading BTCUSDT share one upstream WebSocket connection rather than opening 100 separate streams-a critical efficiency for scaling.
Pros and Cons
Pros
- Open-source with free self-hosted deployment option
- No-code visual strategy builder lowers barrier to entry
- Enterprise-grade multi-tenant architecture suitable for SaaS deployment
- Dual backtesting engines balance speed and accuracy
- AI assistance for strategy generation and analysis
- Efficient market data distribution via Redis fan-out
- Decoupled API/trading engine design prevents trade disruption during updates
Cons
- Self-hosted deployment requires substantial server resources (6+ cores, 16GB RAM)
- Limited exchange support currently (Binance and Bybit are stable; others in development)
- Managed cloud pricing not transparently disclosed
- Complex distributed architecture may present operational challenges for non-technical users
- Limited third-party integrations beyond exchange connections
FAQ
What is DepthSight and how does it differ from other trading bots?
DepthSight is an enterprise-grade, open-source platform for algorithmic cryptocurrency trading. Unlike simple bot frameworks, it provides a complete SaaS-in-a-box architecture with multi-tenant capabilities, visual strategy building, AI assistance, and dual backtesting engines. The platform targets users who want professional-grade automation without writing Python code.
Who should use DepthSight?
The platform appeals to several user types: discretionary traders wanting to systematize their strategies, active scalpers needing real-time market alerts, quantitative traders requiring backtesting and optimization tools, and developers or entrepreneurs looking to launch commercial trading platforms. The self-hosted option suits technically capable users, while the managed cloud targets those preferring turnkey solutions.
What exchanges does DepthSight support?
Currently, Binance and Bybit have full native support and are described as stable. Bitget, OKX, Gate.io, and BingX are in active development. All exchange communication uses CCXT-based executors behind a unified abstraction layer, simplifying the addition of new exchanges through adapter implementation.
Is DepthSight free to use?
The self-hosted deployment runs entirely free on your own infrastructure. You pay only for your VPS costs. A managed cloud option exists for those preferring not to self-host, though specific pricing requires direct inquiry. The open-source codebase is available on GitHub.
What are the system requirements for self-hosting?
DepthSight recommends Ubuntu 22.04+ servers with a minimum of 6 CPU cores and 16 GB RAM. Deployment uses a single shell script that handles Docker installation, secret generation, firewall configuration, and service startup. The platform includes an auto-update mechanism that enables zero-downtime updates without granting root privileges to containers.
How does the AI co-pilot work?
The AI assistant can generate trading strategies from text descriptions or chart screenshots. It analyzes existing strategies for risk factors (suggesting stop-losses for high-volatility scenarios, for example) and reviews live trades and backtest results. The underlying model uses Google Gemini.
What should users verify before relying on DepthSight for live trading?
Users should thoroughly backtest strategies using both the fast vector engine and the detailed candle/tick-level simulator before deploying real capital. Risk management settings require careful configuration. The platform supports paper trading, which users should utilize extensively. Additionally, verify exchange API permissions and understand the security implications of storing API keys within the platform.
Conclusion
DepthSight presents a viable option for traders seeking to automate cryptocurrency strategies without coding expertise, while also providing the architectural foundation for entrepreneurs to build commercial trading platforms. The open-source nature and free self-hosted deployment remove cost barriers, though users should factor in server requirements and operational complexity.
The platform's strength lies in its visual strategy builder, dual backtesting approach, and enterprise-grade multi-tenant architecture. Prospective users should evaluate whether the current exchange support meets their needs and plan for thorough testing before live deployment. Visit depthsight.pro for documentation and deployment instructions.
