The EdgeBase Vision
Future-Ready Fully Managed Backend Platform
Why EdgeBase Exists
Core Insight
Frontend engineers want to build production-grade apps—fast—without dealing with fragmented backend infrastructure, poor observability, or edge-incompatible platforms. But today's backend options are either too generic or too AI-specific, lacking balance and modern design.
Too Many Services
Multiple services stitched manually — high cognitive load, messy integration, and complex setup.
Latency Issues
Cold starts, regional lag — bad UX, especially for global users.
No Real Memory Model
No persistent state per user/session — requires hacks via Redis or local state.
Poor Observability
No tracing, failure, or cost visibility — difficult to debug or optimize.
The Platform
Auth
Email, OAuth, JWT out-of-the-box
PostgreSQL
Global-ready SQL via Hyperdrive or native connection
File Storage
R2-backed object storage with zero egress costs
Functions
Global compute via Workers, zero cold starts
Memory
Per-user/session scoped durable memory via Durable Objects
Knowledge Store
Built-in document ingestion, embedding, querying
Tool Binding
One config to expose HTTP, DB, or external workflows
Observability
Tracing, cost, usage, failure introspection
Developer Experience
Designed for modern product engineers, not backend specialists.
Interfaces
- JS/TS SDK (@edgebase/client) – Works in browser, Node.js, and edge
- CLI – edgebase init, edgebase dev, edgebase deploy, edgebase trace
- Dashboard – Logs, memory browser, usage metrics, deployment view
- Config-as-Code – YAML or edgebase.config.ts
DX Principles
- No boilerplate: Instant setup, zero fetch plumbing
- CLI-first: Local dev, mock sessions, tracing built-in
- Edge-native: Built for latency-sensitive apps
- Fully Managed: No Docker, no cluster config, no glue code
Architecture Overview
Compute
Workers
Low-latency global execution
State
Durable Objects
Session-aware memory, coordination
Storage
R2
Secure object storage
Database
Postgres
Via Hyperdrive or BYODB
Use Cases
1. Fullstack SaaS Backend
Auth, DB, storage, file handling, user sessions. No cloud provider setup required, deploy from CLI.
2. Real-Time or Edge Apps
Global latency guarantees (<60ms). Durable state per user/session.
3. AI-Enhanced Products
Tools, memory, observability + gateway support. Declarative config, no additional frameworks required.
What We're Not
Not ML-ops tooling — We're focused on backend infrastructure, not ML model training or deployment pipelines.
Not CRUD-only — While we support CRUD out-of-box, we're built for complex, real-time, edge-native applications.
Not glued-together infra — Everything is unified, not a collection of services you need to wire together.
Not opinionated about frontend — Works with any frontend stack, framework, or platform.
We're the future-ready backend, engineered for modern product teams.
Strategic Positioning
EdgeBase is a fully managed backend for modern apps — fast to build, scalable by design, and built on edge-native primitives. It replaces 5+ tools and lets you ship faster, debug easier, and scale cheaper.
Edge-native
Speed, reliability, no cold starts
Modern dev ergonomics
SDK + CLI + dashboard
Optional AI stack
Memory, tools, gateway, observability
Unified DX
Fullstack without glue code or duct tape
Roadmap Concepts
Immediate
- Core SDK (auth, memory, tools)
- CLI + config setup
- Postgres & R2 integration
- Edge-native deploy flow
Mid-Term
- Observability: trace viewer, token cost, errors
- Knowledge store (docs, embeddings, queries)
- Secure rate limiting + metering
- Gateway support for multiple AI models and OSS
Long-Term
- Multi-agent orchestration (optional)
- Private model hosting gateway
- Org/team usage controls + billing