Senren
An ML platform for managing infrastructure across clouds, regions, and clusters.
Senren provides a unified control plane for deploying and managing ML infrastructure—feature stores, search indexes, model servers, and databases—across AWS, GCP, and multiple regions.
from senren import SenrenClient, Database
client = SenrenClient(endpoint="control.senren.dev")
# Deploy feature store across AWS + GCP
client.apply_state(
databases=[
Database(
name="user-features",
type="redis-cluster",
memory_mb=8192,
regions=[
"aws:us-east-1",
"gcp:us-central1",
"gcp:europe-west1",
],
)
]
)
Senren automatically provisions infrastructure in each region, synchronizes state, and reports status back.
Why Senren?
Multi-cloud native Deploy to AWS, GCP (Azure coming) from a single Python API.
Declarative state management Define desired state, Senren reconciles across all clusters.
Regional autonomy Regional planes operate independently—no central bottleneck.
Built for ML workloads Feature stores, search indexes, model serving, not just databases.
Learn more about the motivation →
How It Works
┌─────────────────┐
│ Python SDK │ Your code
└────────┬────────┘
│ gRPC
┌────────▼────────┐
│ Control Plane │ Stores desired state
│ (PostgreSQL) │
└────────┬────────┘
│ Kafka
┌────┴────┬────────┬────────┐
▼ ▼ ▼ ▼
┌───────┐ ┌───────┐ ┌───────┐ ┌───────┐
│AWS │ │AWS │ │GCP │ │GCP │
│us-e-1 │ │eu-w-1 │ │us-c-1 │ │eu-w-1 │
│ │ │ │ │ │ │ │
│Regional│ Regional Regional Regional │
│Plane │ │Plane │ │Plane │ │Plane │
└───┬───┘ └───┬───┘ └───┬───┘ └───┬───┘
│ │ │ │
▼ ▼ ▼ ▼
K8s API K8s API K8s API K8s API
- Define: Use Python SDK to declare desired infrastructure
- Sync: Control plane publishes state to Kafka
- Apply: Regional planes compute diffs and apply to Kubernetes
- Provision: Controllers provision actual infrastructure
- Monitor: Status flows back through Kafka to control plane
Current Status
MVP (Available Now): - ✅ Multi-region database management (Redis, Redis Cluster, Dragonfly DB) - ✅ Multi-cloud support (AWS, GCP) - ✅ State-based synchronization via Kafka - ✅ Python SDK
Roadmap: - 🚧 Feature stores (online/offline) - 🚧 Search indexes (Elasticsearch, OpenSearch) and document stores - 🚧 Vector databases for semantic search - 🚧 Model serving infrastructure - 🚧 A/B testing with automatic traffic routing
Get Started
Deploy your first Redis database in 5 minutes:
Installation → | Quick Start →
Documentation
Overview - Why Senren? - Problems and solutions - Architecture - How Senren works - Roadmap - Current status and future plans
Getting Started - Installation - Quick Start - Core Concepts
Infrastructure (Current MVP) - Databases - Redis, Dragonfly DB - Multi-cloud Setup - AWS + GCP
API Reference - Python SDK