Skip to main content

Getting Started with sypho

sypho is a managed platform for deploying and running AI agents. Build once, deploy anywhere, scale automatically.

What is sypho?​

sypho lets you:

  • Build AI agents with Python and simple decorators
  • Deploy instantly to our managed infrastructure
  • Scale automatically across dev, staging, and production
  • Access knowledge through structured data APIs
  • Monitor everything with built-in logging and debugging

No DevOps, no infrastructure management, no complexity.

Quick Start​

1. Install the CLI​

curl -sSL https://install.sypho.dev | sh

2. Sign up and authenticate​

sypho login
# Opens app.sypho.ai to get your token

3. Create your first agent​

sypho init my-agent
cd my-agent

# Build the agent image
sypho build --version 0.1.0

# Deploy to sypho
sypho deploy --version 0.1.0

# Run it
sypho run my-agent:0.1.0 --input '{"message": "Hello!"}'

That's it. Your agent is live on sypho's infrastructure.

How It Works​

  1. Write agents using Python and the sypho SDK
  2. Build with the CLI (generates manifest, builds Docker image)
  3. Deploy to app.sypho.ai (registers your agent)
  4. Run via CLI or API (executes on our infrastructure)

Key Concepts​

Agents​

Python applications with:

  • Entrypoints - Functions that define what your agent does
  • Tools - Functions the LLM can call
  • Agent Config - LLM model, system prompt, iteration limits
from sypho_sdk import entrypoint, tool, AgentContext

@tool
async def fetch_data(query: str):
    """Fetch data from an API"""
    return {"result": "..."}

@entrypoint
async def main(input: dict, context: AgentContext):
    """
    You are a data assistant. Help users query data.
    """
    pass  # LLM decides which tools to call

Deployments​

Organize your agents by environment:

  • dev - Development and testing
  • staging - Pre-production validation
  • prod - Production workloads

Runs​

Each execution of your agent:

  • Created via CLI or API
  • Tracked with logs and events
  • Billed per execution

Platform Features​

Knowledge Management​

Store and query structured data:

# Save data
await context.call_tool("structured_data_save", {
    "namespace": "my-data",
    "key": "user:123",
    "data": {"name": "Alice"}
})

# Query data
results = await context.call_tool("structured_data_query", {
    "namespace": "my-data",
    "filters": {"metadata.type": "user"}
})

Built-in Tools​

Every agent gets access to:

  • Structured data storage
  • Query capabilities
  • Data versioning

Monitoring​

  • Real-time logs via CLI: sypho logs <run_id>
  • Event tracking in dashboard
  • Usage metrics and billing

Pricing​

Pay only for what you use:

  • Free tier - 100 runs/month
  • Pro - $0.01 per run
  • Enterprise - Custom pricing

Next Steps​