EdgeDB Cloud: The Database for the Edge Era

The Problem

The modern web is global, but databases are still stuck in one place.

Your application might be deployed to the edge — on Cloudflare Workers, Vercel Edge Functions, or Deno Deploy — running in 300+ locations worldwide. But your database? It’s still sitting in us-east-1.

Every request from Tokyo, São Paulo, or Mumbai pays a 200–400ms round trip tax to Virginia before a single byte of data reaches the user.

This latency tax is devastating:

• Amazon found that every 100ms of latency costs 1% in sales
• Google discovered that a 500ms delay reduces traffic by 20%
• User patience is finite — 53% of mobile users abandon sites that take more than 3 seconds to load

Existing solutions try to patch around this, but all have critical flaws:

• Read replicas (e.g., PlanetScale, CockroachDB)

  • Reduce read latency, but writes still funnel to a primary region
  • Create inconsistency windows, complex conflict resolution, and operational overhead

• CDN caching

  • Great for static content
  • Falls apart for personalized, authenticated, or real-time data where cache correctness is hard

• In-memory caches (Redis, Memcached)

  • Add another layer to manage
  • Introduce cache invalidation complexity — one of the classic hard problems in computer science

• SQLite at the edge (Turso, LiteFS)

  • Promising for some workloads
  • But constrained to single-writer architectures, small datasets, and tricky replication models

The fundamental issue: no mainstream database has been designed from first principles to be globally distributed, with local-speed reads and writes, for edge-native architectures.

The Solution: EdgeDB Cloud

EdgeDB Cloud is a globally distributed, edge-native database purpose-built for the edge era.

It delivers sub-10ms reads from anywhere in the world by automatically replicating your data to 50+ edge locations, while maintaining strong, tunable consistency guarantees for writes.

In a world where compute runs everywhere, EdgeDB Cloud makes your data follow your users — automatically.

How It Works

1. Create a database in one command

   • Define your schema
   • Choose your regions (or let EdgeDB auto-select based on user distribution)
   • Get a globally distributed database in under 60 seconds

2. Data goes where your users are

   • EdgeDB Cloud automatically replicates your data to edge locations closest to your users
   • A read query from Tokyo hits a local replica, not a server 6,000 miles away
   • Result: single-digit millisecond reads globally

3. Smart write routing

   • Writes are automatically routed to the optimal region via a proprietary consensus protocol
   • Cross-region write propagation completes in under 50ms
   • Consistency is tunable per query:
     • Strong — for money, inventory, critical state
     • Causal — for user sessions, feeds, collaborative apps
     • Eventual — for analytics, logs, non-critical counters

4. Schema-aware optimization

   • EdgeDB Cloud is schema-aware and understands your data model and access patterns
   • Hot data is proactively pre-positioned in more regions
   • Cold data is kept in fewer regions to minimize storage costs
   • You get global performance without manually tuning replication topologies

Key Technical Innovations

1. Adaptive Topology

EdgeDB Cloud uses an adaptive replication topology that self-optimizes based on real traffic.

• If 80% of reads come from Europe, EdgeDB Cloud automatically increases European replication density
• As your user base shifts, the topology rebalances itself — no manual reconfiguration, no downtime
• You get the best of both worlds: low latency for active regions and cost efficiency for inactive ones

2. Conflict-Free Writes with CRDTs

Not all data needs strict serializability.

For operations where strong consistency isn’t required — like counters, append-only logs, activity feeds, and social reactions — EdgeDB Cloud uses CRDTs (Conflict-free Replicated Data Types).

• Enables true multi-region writes
• Zero conflicts by construction
• No manual conflict resolution logic in your application

You choose per operation whether you want strict consistency or highly available, conflict-free writes.

3. Time-Travel Queries

EdgeDB Cloud lets you query your data at any point in the last 30 days.

This unlocks:

• Debugging: “What did this record look like before the bug?”
• Auditing & compliance: Full historical visibility
• Forensics: “What changed, when, and by whom?”

Time-travel queries turn your database into a built-in audit log and debugging tool, without extra infrastructure.

4. Edge-Native Query Engine

The EdgeDB Cloud query engine is edge-native:

• Compiles to WebAssembly
• Runs directly on edge runtimes (e.g., Cloudflare Workers, Vercel Edge Functions)
• Query planning happens locally, near your code

This minimizes round trips to the core cluster and lets your application:

• Parse, validate, and plan queries at the edge
• Send optimized, compact requests to the data plane
• Achieve end-to-end latency that’s often faster than in-region traditional databases

Traction

EdgeDB Cloud is not a prototype — it’s already powering production workloads.

• 3,000+ databases deployed across startups, scale-ups, and enterprises
• 99.999% uptime over the last 18 months — five nines of reliability
• $8M Series A led by Andreessen Horowitz (a16z), with participation from Elad Gil and the founders of Vercel and Cloudflare
• 50+ edge locations across 6 continents
• Global P99 read latency: 8ms — faster than many in-region database queries
• 400+ production applications in gaming, e-commerce, fintech, and SaaS

This combination of performance, reliability, and backing positions EdgeDB Cloud as a leading contender in the emerging edge-native database category.

Business Model

EdgeDB Cloud uses a usage-based, pay-per-query model that scales from side projects to global platforms.

• Free Tier

  • 1 database
  • 3 edge regions
  • 1GB storage
  • 100K queries/day
  • Generous enough for side projects, MVPs, and early-stage experiments

• Pro — $29/month + usage

  • Unlimited regions
  • 50GB included storage
  • Auto-scaling
  • Point-in-time recovery
  • Branching for preview environments and safe schema evolution

• Enterprise — $499/month + usage

  • Dedicated infrastructure
  • VPC peering
  • Custom regions
  • SLA-backed performance and availability
  • SOC 2 compliance
  • Priority support and onboarding

The pay-per-query pricing ensures you:

• Start at $0
• Scale to millions or billions of queries
• Avoid manual provisioning and capacity planning

Market Opportunity

The cloud database market is currently valued at $22B, growing at 18% annually.

But the shift to edge computing is creating a new, high-growth subcategory: edge-native databases.

• This segment is projected to reach $4B by 2028
• As more applications adopt edge architectures, the need for globally distributed, low-latency data becomes critical

Most incumbents are retrofitting traditional databases for the edge — bolting on read replicas, caches, and regional hacks.

EdgeDB Cloud is different:

• Designed from the ground up for a world where compute is everywhere
• Built to make data follow users automatically, not as an afterthought
• Positioned as the default data layer for edge-native applications

Team

EdgeDB Cloud is built by a team that has already solved these problems — at scale.

• CEO

  • Former Principal Engineer at CockroachDB
  • Led the geo-distributed replication team
  • Deep expertise in distributed SQL and multi-region consistency

• CTO

  • Former Cloudflare Workers architect
  • Helped design the edge computing platform that serves 20% of the internet’s traffic
  • Brings a first-principles understanding of edge runtimes and global infrastructure

• Head of Research

  • PhD in Database Systems from MIT
  • Specializes in distributed consensus protocols and fault-tolerant systems

This team combines:

• Database internals expertise (CockroachDB)
• Edge platform experience (Cloudflare Workers)
• Cutting-edge research (MIT, consensus, CRDTs)

EdgeDB Cloud is the synthesis of these perspectives — a database that is natively global, edge-aware, and developer-friendly.

Why Now

• Edge platforms (Cloudflare Workers, Vercel, Deno Deploy) are maturing rapidly
• Developers are already building at the edge, but are blocked by legacy, region-bound databases
• User expectations for speed and responsiveness are higher than ever

There is a clear gap: a globally distributed, strongly consistent, edge-native database that just works.

EdgeDB Cloud fills that gap.

Vision

EdgeDB Cloud makes distance irrelevant.

In a world where compute runs everywhere, your data should too. With EdgeDB Cloud:

• Reads are local
• Writes are fast and consistent
• Topology is adaptive
• Complexity is abstracted away

Developers stop fighting latency and replication, and start focusing on what matters: building great products for a global user base.