Choosing Your Approach
Datafye offers four different deployment scenarios, each designed for specific use cases and levels of control. This guide helps you understand the differences and choose the right approach for your needs.
Quick Start: If you already know which scenario you want, jump to Choose Your Path to get started immediately.
The Four Scenarios
Foundry: Data Cloud Only
The Data Cloud Only scenario is designed for developers who already have their own algorithmic trading infrastructure and primarily need access to high-quality, normalized market data. In this scenario, Datafye provides you with comprehensive access to both historical and real-time market data through REST and WebSocket APIs. You retain complete control over your algo implementation by running your own containerized algo engines, bringing your own backtesting infrastructure if needed, and using your own optimization tools if desired.
This scenario is best suited for developers with existing algo infrastructure who want better data quality, teams with custom backtesting frameworks that they prefer to keep using, users who want maximum control over every aspect of their algo implementation, and cost-conscious developers who already have the tools they need and just want to add Datafye's data to their workflow.
Foundry: Full Stack
The Full Stack Foundry scenario provides a complete algo development environment. You get everything from the Data Cloud Only scenario plus the Datafye Algo Container SDK for building algos, AI-assisted vibe coding through MCP server integration that accelerates development, a high-performance backtesting engine with parallelization capabilities, genetic algorithm-based parameter optimization to find the best strategy configurations, and comprehensive performance scorecarding to evaluate your strategies.
In this scenario, you focus on providing your algo logic built using the SDK and your strategy ideas and parameters to optimize, while Datafye handles all the infrastructure. This approach is ideal for developers who want the complete algo development experience without building infrastructure, teams who want to leverage AI-assisted vibe coding for rapid iteration and experimentation, users who need sophisticated backtesting and optimization capabilities, and developers building algos from scratch without existing tools.
Trading: Data Cloud + Broker
The Trading: Data Cloud + Broker scenario bridges the gap between development and live trading for teams with existing algo engines. Datafye provides live market data feeds, a broker connector for both paper and live trading, portfolio and position management capabilities, and robust trade execution infrastructure. You continue to run your own containerized algo engines that generate trading signals.
This scenario works best for developers with existing algos who are ready to start trading them, teams with custom algo infrastructure who need execution capabilities, users who want to maintain complete control over their proprietary algo logic, and developers transitioning from backtesting environments to live trading without rewriting their entire codebase.
Trading: Full Stack
The Trading: Full Stack scenario provides the most comprehensive end-to-end solution. You receive live market data feeds, the Datafye Algo Container for algo execution, AI-assisted vibe coding that works even in production for live iteration, a broker connector for paper and live trading, portfolio and position management, and real-time monitoring and metrics.
This is the complete solution for teams wanting integrated data, execution, and monitoring in one package. It's perfect for users who want Datafye to handle all infrastructure concerns, teams leveraging AI-assisted vibe coding not just in development but also in production for quick adjustments, and developers who prefer a turnkey solution over assembling separate components.
Comparison Matrix
Market Data
✅
✅
✅
✅
Algo Container SDK
❌
✅
❌
✅
AI-Assisted Vibe Coding
❌
✅
❌
✅
Backtesting Engine
❌
✅
❌
❌
Genetic Optimization
❌
✅
❌
❌
Scorecarding
❌
✅
❌
❌
Broker Connector
❌
❌
✅
✅
Paper Trading
❌
❌
✅
✅
Live Trading
❌
❌
✅
✅
Your Own Containers
✅
❌
✅
❌
Full Control
⭐⭐⭐
⭐⭐
⭐⭐⭐
⭐⭐
Integrated Experience
⭐
⭐⭐⭐
⭐⭐
⭐⭐⭐
Finding Your Scenario
If Your Primary Goal is Better Market Data
You probably want Data Cloud Only if you already have existing algo infrastructure, just need high-quality normalized data to improve your current system, and want to integrate Datafye data into your existing workflow without changing your development process. This scenario lets you keep everything else you've built while upgrading your data quality.
If Your Primary Goal is Developing and Optimizing New Algos
You probably want Full Stack Foundry if you're building new trading strategies from scratch, want sophisticated backtesting and optimization capabilities to validate and improve your ideas, want to leverage AI-assisted vibe coding to accelerate your development cycle, and need performance scorecarding to objectively evaluate strategy performance before deploying capital.
If Your Primary Goal is Trading Existing Algos
You probably want Data Cloud + Broker if you have working algo containers that you've already developed and tested, need live data and execution infrastructure to trade those algos, want to start paper or live trading without rewriting your code, and want to maintain complete control over your proprietary algo implementation and intellectual property.
If Your Primary Goal is a Complete Algo Trading Solution
You probably want Full Stack Trading if you want data, algos, and execution integrated in one cohesive package, are building new trading strategies and want to go from idea to live trading within the same platform, want integrated monitoring and management of all aspects of your trading, and want to leverage AI-assisted vibe coding not just during development but also in production for rapid adjustments.
Real-World Use Cases
Use Case 1: The Quantitative Researcher
Let's say you're a quant researcher with Python-based research tools and custom backtesting infrastructure that you've spent years perfecting. You have your own analysis workflows, your team knows your internal tools inside and out, and you're not looking to change your development process—you just need better quality market data to improve your research output.
For this situation, Data Cloud Only is the right choice. You already have analysis and backtesting tools that work well for you, you just need high-quality historical data to feed into your existing systems, you want to integrate Datafye data into your existing workflow with minimal disruption, and you don't need Datafye's backtesting or execution capabilities because you've already built those yourself. Your next step would be the Foundry: Data Cloud Only Quickstart.
Use Case 2: The Individual Algo Developer
Imagine you're an individual developer learning algo trading and want to build your first strategy. You're excited about the field but don't have existing infrastructure, and you're not sure where to start with backtesting and optimization. You want guidance and tools that help you learn while building real strategies.
For your situation, Full Stack Foundry makes the most sense. You don't have existing infrastructure so you're not locked into any particular toolset, you want guidance through AI-assisted vibe coding that can help you learn best practices as you build, you need backtesting capabilities to validate your ideas before risking real money, you want optimization tools to find the best parameters for your strategy, and you're not ready to trade yet—you want to learn and build confidence first. Start with the Foundry: Full Stack Quickstart.
Use Case 3: The Prop Trading Firm
Consider a proprietary trading firm that has invested heavily in building their own algo engines over several years. These engines represent significant intellectual property and competitive advantage. The firm wants to deploy these algos with better data and execution capabilities, but they don't want to rewrite their core logic or share it with external systems.
This firm should choose Data Cloud + Broker. They have proven algo infrastructure that represents their intellectual property and competitive advantage, they need low-latency market data to improve their execution quality, they want to trade their algos with minimal changes to their existing codebase, and they want to maintain complete control over their proprietary algo logic. Their next step is the Trading: Data Cloud + Broker Quickstart.
Use Case 4: The Aspiring Marketplace Publisher
Think about a developer who wants to build an algo to publish on the Datafye Marketplace, where other users can subscribe to and trade their strategy. This developer needs to not only build and validate the algo but also demonstrate its performance through rigorous testing and scorecarding.
This developer should start with Full Stack Foundry, then move to Full Stack Trading. They need to develop and optimize the algo first using Foundry's tools, they need scorecarding for marketplace submission to prove the strategy's performance, they need to validate with paper trading using the Trading environment before taking it live, and the marketplace requires algos built with the Datafye SDK so they must use Full Stack scenarios. The path forward is: Foundry: Full Stack Quickstart, then Trading: Full Stack Quickstart, and finally Publishing to Marketplace.
Use Case 5: The Hedge Fund Manager
A hedge fund manager wants to deploy multiple strategies with institutional-grade infrastructure. The fund has various teams with different needs: some teams have existing infrastructure they want to preserve, others are developing new strategies, and some are ready to deploy to production. The manager needs flexibility to support all these teams.
For this situation, the recommendation depends on each team's specific needs. Teams with existing infrastructure who just need better data should use Data Cloud Only. Teams developing new strategies who want optimization tools should use Full Stack Foundry. Teams deploying existing containers to production should use Data Cloud + Broker. And teams wanting the complete integrated solution should use Full Stack Trading. The key insight here is that you can run multiple deployments with different scenarios for different teams or strategies, allowing each team to work in the way that suits them best. The next step is to contact Datafye for an enterprise consultation to discuss the specific needs of each team.
Migration Paths
You're not locked into one scenario forever. Many users start with one approach and migrate to another as their needs evolve. Here are common migration paths:
From Data to Development to Trading
Many teams start with Data Cloud Only to integrate Datafye data into their existing systems, validate that the data quality meets their needs, and build initial algos using their own tools. As they become more comfortable with the platform, they add Full Stack Foundry to migrate their algos to the Datafye SDK, leverage backtesting and optimization capabilities they didn't have before, and use AI-assisted vibe coding for rapid iteration on new ideas. Finally, when algos are validated and ready for production, they deploy using Full Stack Trading to paper trade validated algos first, gradually transition to live trading with real capital, and monitor performance using integrated tooling.
From Development to Trading
A simpler path for new developers is to start with Full Stack Foundry where they build and optimize algos from scratch, backtest thoroughly using historical data, and generate scorecards that demonstrate performance. When ready, they move to Full Stack Trading to paper trade first in simulated conditions, validate strategy behavior in real market conditions, and go live when they're confident in the strategy's performance.
From Own Containers to Datafye SDK
Some teams start with Data Cloud Only, using their existing containers with Datafye data to evaluate the platform without committing to a full migration. As they see the benefits, they migrate to Full Stack Foundry where they rewrite algos using the Datafye SDK to take advantage of integrated tooling, leverage AI-assisted vibe coding and optimization capabilities, and compare performance against their original implementation. Finally, they trade using Full Stack Trading to deploy SDK-based algos in production and use the integrated execution infrastructure.
Deployment Models
Once you've chosen your scenario (Data Cloud Only, Full Stack Foundry, etc.), you need to decide how to deploy it. Datafye supports two fundamentally different deployment models, each with different characteristics and use cases.
Standalone Deployment
In a Standalone deployment, all Datafye services run on a single machine. This machine can be your desktop, laptop, a physical server, a virtual server, or even a cloud IaaS instance—as long as it has sufficient resources to run all the services.
Key characteristics:
All services (data cloud, algo container, broker connector, etc.) run on one machine
Simple setup and management
Lower resource utilization when idle
Free to use (no infrastructure costs beyond your own machine)
Available for:
Algo development (all Foundry scenarios)
Paper trading only (not live trading)
Best for:
Learning and experimenting with Datafye
Development and testing of new strategies
Running on your local machine during development
Paper trading validation before moving to production
Requirements:
Sufficient resources on the machine (see Prerequisites in quickstart guides)
Docker installed and running
Distributed Deployment
In a Distributed deployment, Datafye services are spread across multiple machines, providing scalability, reliability, and production-grade infrastructure. This is the deployment model for serious development work and all production trading.
Key characteristics:
Services distributed across multiple machines for redundancy
Higher availability and reliability
Better performance for compute-intensive workloads
Scalable based on your needs
Where you can deploy:
IaaS cloud providers (AWS, Azure, GCP, ...)
Private physical machine banks (coming soon)
Distributed deployments come in two flavors:
Self-managed Distributed
You deploy in your own cloud account and manage the infrastructure yourself.
Available for:
Algo development (all Foundry scenarios)
Paper trading only (not live trading)
Best for:
Teams with cloud expertise who want full control
Organizations with existing cloud infrastructure
Development and testing at scale
Paper trading validation with production-like infrastructure
Cost: Free (you pay only your cloud provider's infrastructure costs)
Datafye Managed Distributed
Datafye deploys in a cloud account dedicated to you under the Datafye root account and operates the infrastructure as a managed service.
Available for:
Algo development (all Foundry scenarios)
Paper trading
Live trading (only available in this deployment option)
Best for:
Production live trading
Teams who want Datafye to handle all operational concerns
Organizations requiring compliance, monitoring, and operational excellence
Publishing to the Datafye Marketplace
Cost: Paid managed service
Important: Live trading is only available in Datafye Managed Distributed deployments. This ensures proper compliance, monitoring, risk management, and operational excellence for production trading with real capital.
Choosing Your Deployment Model
The decision tree is straightforward:
Are you live trading with real money?
Yes → You must use Datafye Managed Distributed
No → Continue to next question
Are you paper trading or developing at scale?
Yes, and I want Datafye to manage it → Use Datafye Managed Distributed
Yes, and I want to manage it myself → Use Self-managed Distributed
No, just learning/experimenting → Use Standalone
Do you need production-grade reliability and uptime?
Yes → Use Distributed (Self-managed or Datafye Managed)
No → Standalone is fine
For more details on deployment options, see Private Cloud Deployments.
Cost Considerations
The cost of running Datafye depends on your deployment model choice, not on which scenario (Data Cloud Only, Full Stack, etc.) you select. All four scenarios can be deployed using any of the three deployment models, with costs determined by the deployment model.
Deployment Model Costs
Standalone Deployment: Free
No Datafye infrastructure costs
You provide the machine (desktop, laptop, server, or cloud VM)
You pay only for the machine's costs (electricity for local, cloud VM costs for cloud)
Best for: Learning, development, and paper trading
Distributed Self-managed: Free
No Datafye infrastructure costs
You deploy in your own cloud account (AWS, Azure, GCP, etc.)
You pay only your cloud provider's infrastructure costs (VMs, networking, storage)
Best for: Teams with cloud expertise who want full control and production-like infrastructure without managed service fees
Distributed Datafye Managed: Paid
Managed service with dedicated infrastructure operated by Datafye
Pricing includes infrastructure, monitoring, compliance, and operational support
Required for live trading
Best for: Production live trading and teams wanting Datafye to handle all operational concerns
Scenario Considerations
While costs are determined by deployment model, different scenarios have different resource requirements:
Data Cloud Only requires minimal infrastructure (just data services)
Full Stack Foundry requires more resources for backtesting and optimization engines
Trading scenarios require additional infrastructure for broker connectivity and portfolio management
Full Stack Trading requires the most comprehensive infrastructure stack
These resource differences affect your infrastructure costs in Self-managed deployments (you pay for more cloud resources) and may affect pricing in Datafye Managed deployments.
Optimizing Your Costs
Start with Standalone for:
Learning and experimenting with Datafye
Initial algo development and testing
Paper trading validation before committing to production
Non-time-critical development work
Move to Distributed Self-managed when you need:
Production-grade reliability without managed service costs
Scalability for high-volume backtesting
Your team has cloud expertise to manage infrastructure
Paper trading at scale with production-like infrastructure
Use Distributed Datafye Managed when you need:
Live trading with real capital (required)
Datafye to handle all infrastructure operations
Compliance, monitoring, and operational excellence
To focus on algo development rather than infrastructure management
Publishing to the Datafye Marketplace
Deployment Restrictions
Remember these important restrictions when planning your deployment:
Standalone: Supports algo development and paper trading only (not live trading)
Distributed Self-managed: Supports algo development and paper trading only (not live trading)
Distributed Datafye Managed: Supports algo development, paper trading, AND live trading
For live trading, you must use Distributed Datafye Managed deployment, which ensures compliance, monitoring, and operational excellence.
Making the Final Decision
If you're still not sure which scenario is right for you, consider these questions:
First, do you have existing algo infrastructure? If yes, you probably want Data Cloud Only or Data Cloud + Broker to leverage your existing investment. If no, you probably want Full Stack Foundry or Full Stack Trading to get started quickly without building infrastructure.
Second, are you ready to trade? If yes, you want Data Cloud + Broker or Full Stack Trading to access execution capabilities. If no, you want Data Cloud Only or Full Stack Foundry to focus on development and validation first.
Third, do you want Datafye's backtesting and optimization? If yes, you need Full Stack Foundry to access these capabilities. If no, Data Cloud Only may be sufficient.
Fourth, do you want to use AI-assisted vibe coding? If yes, you need Full Stack scenarios where this is available. If no, Data Cloud Only or Data Cloud + Broker will work fine.
Finally, do you plan to publish to the Marketplace? If yes, you must use Full Stack scenarios as the Marketplace requires SDK-based algos. If no, any scenario will work based on your other needs.
Next Steps
When you're ready to get started, begin by reviewing the quickstart guide for your chosen scenario: Foundry: Data Cloud Only if you want to integrate Datafye data with your own containers, Foundry: Full Stack if you want the complete development environment, Trading: Data Cloud + Broker if you want to trade your own containers, or Trading: Full Stack if you want the complete trading solution.
You should also understand your deployment options by reading about Private Cloud Deployments to choose between Standalone, Distributed Self-managed, or Distributed Datafye Managed deployments. Learn about configuration by exploring Descriptor Reference to understand how to specify your data, algo, and broker requirements. And get the CLI by following the CLI Installation guide to install the tools you need to provision and manage your deployment.
To learn more about the platform architecture and how these scenarios fit together, see What is Datafye?.
Last updated