Distributed Systems • Deterministic Runtimes • Causal Infrastructure
📥 james@flyingrobots.dev · 🔗 LinkedIn · 📄 Résumé
I've spent two decades building systems where correctness survives concurrency: AAA game engines, MMORPG backends, live-service platforms, high-throughput distributed infrastructure.
Across all of it, the same failure mode kept surfacing: state forgets how it became itself.
Race conditions, irreproducible bugs, and partial failures aren’t random; they’re what happens when history is invisible and mutation is implicit.
The Continuum Stack is my answer: systems where history is explicit, replay is exact, and behavior is derived, not implied.
- Continuum — protocol for distributed systems built on causal history instead of shared mutable state
- Echo — deterministic runtime where every tick is witnessed and fully reproducible
- Wesley — schema compiler that generates enforceable runtime contracts, codecs, and versioned causal semantics
- git-warp — Git-native causal storage for offline-first systems with strong provenance
- WARP TTD — time-travel debugger that follows causal chains, not log lines
- WARPDrive (in development) — FUSE layer that materializes causal history as ordinary files and directories
- Distributed coordination, replication, and convergence
- Deterministic runtimes and simulation infrastructure
- Observability, replayability, and causal debugging
- Offline-first and synchronization-heavy architectures
- High-stakes correctness and auditability
I’d love to talk if you've got problems that need correctness after deployment, not just before it.