Backend & Distributed Software Engineer

I build backend systems for the regime where one machine stops being an option — where partitioning, throughput, and failure handling are design constraints rather than afterthoughts.

My work centers on distributed data processing at scale: streaming ingestion over Kafka, multi-threaded transformation engines, and wide-column stores like ScyllaDB, where the access pattern has to be designed around the partition key instead of discovered in production. Most of the interesting engineering lives in the unglamorous parts — shard-aware batching, backpressure, idempotent retries, and choosing a consistency model you can actually defend.

I reach for event-driven architecture for the same reason I reach for clean interfaces: it lets services fail independently. My design instinct goes toward staying correct under concurrency and predictable at the tail, not just fast in the average case.

Lately I've been deep in the cloud-native ecosystem — Kubernetes controllers, operators, and Opensource CNCF projects — and in Agentic AI, building MCP-based tooling that gives models typed, permissioned access to real developer systems.

I'm drawn to low-level systems problems, code that has to hold up in production, and the gap between working and correct.

Languages

Backend & Architecture

Data & Streaming

Cloud & Platform

AI Engineering

• 🧩 Consensus and replication internals — Raft, quorum reads, and what "eventual" actually costs you
• ⚙️ Kubernetes operators — writing controllers instead of clicking through YAML
• 🔍 Query engine internals — vectorized execution and why columnar layouts win
• 🤖 Agentic tooling on MCP — giving models safe, typed access to real developer systems

Always happy to talk about distributed systems.