I am a backend-first engineer focused on building systems that stay understandable under pressure: explicit boundaries, predictable failure modes, memory-conscious design, and tooling that makes complex work feel smaller.
- 🔭 Current Focus: Distributed systems, .NET internals, idempotency, orchestration, and Model Context Protocol servers.
- 💡 Core Competency: Turning messy domains into typed, composable backend primitives.
- 🌱 Learning Edge: Runtime behavior, Roslyn-powered automation, search/indexing algorithms, and AI-assisted developer workflows.
🧭 convoy
Multi-tenant orchestration for .NET.
A backend-first orchestration project shaped around tenant-aware execution, operational clarity, and clean service boundaries.
- Focus: Multi-tenancy, orchestration, system design, .NET services.
- Why it matters: It explores the kind of infrastructure layer that keeps complex business workflows controlled and observable.
🪄 Levitate
Compile C# automation methods into production-grade PowerShell scripts with Roslyn.
A developer automation project that bridges strongly typed C# authoring with practical script output.
- Focus: Roslyn, automation, C#, PowerShell generation.
- Why it matters: It treats scripts as build artifacts, not throwaway glue.
Idempotency primitives for reliable .NET APIs.
A fault-tolerance project for making retries, duplicate requests, and distributed edge cases easier to reason about.
- Focus: Idempotency keys, API reliability, duplicate request handling, distributed systems.
- Why it matters: Reliable retries are boring only when the system gets them right.
🗒️ Mcp.Obsidian
An MCP server for connecting AI agents with Obsidian knowledge workflows.
An MCP server for SQL Server, focused on giving agents a safer bridge into relational data workflows.
A Rust package for faster grep operations.
Built around fast exact search plus native edit-distance search, inspired by research-paper-driven search/indexing ideas.
🧩 UBHeap
Partial-order sorting and heap primitives for modern .NET.
Inspired by unified-bound heaps research, built for dependency graphs, schedulers, workflow engines, and repeated DAG sorting.
A high-performance C#/.NET priority queue inspired by the 2026 SimdQuickHeap paper.
Built around SIMD-aware heap layouts for large task schedulers, graph workloads, simulations, and interleaved push/pop queues.

