Backend & Distributed Systems Engineer
I build systems that handle scale — distributed auth, fault-tolerant streaming pipelines, and high-throughput microservices.
I work at the intersection of distributed systems, backend infrastructure, and platform engineering — designing services that are fast, fault-tolerant, and observable at scale.
- Distributed systems — consensus protocols (Raft), gossip-based membership, sharded architectures
- High-throughput pipelines — event streaming (Kafka, Pub/Sub), batch processing (Spark, Delta Lake), real-time analytics
- Backend services — microservices in Go/Java, REST + gRPC APIs, adaptive caching, query optimization
- Cloud-native infra — Kubernetes, Docker, OCI, GCP, CI/CD, GitOps
- AI-powered tooling — LLM-based automation, vector search, RAG pipelines, observability sidecars
| Area | Technologies |
|---|---|
| Languages | Go · Java · Python · SQL · TypeScript |
| Backend & Systems | Distributed Systems · Microservices · REST · gRPC · Concurrency |
| Streaming & Data | Apache Kafka · Apache Spark · Pub/Sub · BigQuery · Delta Lake |
| Databases | PostgreSQL · Redis · Oracle ATP · MongoDB |
| Cloud & Infra | Kubernetes · Docker · OCI · GCP · Terraform · CI/CD |
| AI & Search | LLMs · Vector DBs · HuggingFace · RAG · Embeddings |
Sharded, fault-tolerant key-value store built from scratch in Go
- Implements Raft consensus for leader election and log replication across nodes
- Uses gRPC for inter-node RPC communication
- Gossip protocol for cluster membership and failure detection
- Designed to survive node failures without data loss or split-brain
Go gRPC Raft Distributed Systems Consensus
High-throughput URL shortener designed around read-heavy workload patterns
- Bloom filter to eliminate DB hits for non-existent keys
- Redis caching layer for sub-millisecond lookup on hot URLs
- PostgreSQL for persistence with write-optimized schema
- Built to handle high read throughput with minimal write amplification
Go Redis PostgreSQL Bloom Filter System Design
Real-time collaborative coding platform with audio and cloud execution
- WebRTC for peer-to-peer audio communication
- WebSocket for real-time code sync between participants
- Cloud-based code execution sandbox
- Built with Go backend + HTMX frontend
Go HTMX WebRTC WebSocket Real-time Systems
Concurrent bulk email dispatch tool in Go
- Leverages Go goroutines and channels for concurrent dispatch
- Rate-limited SMTP client with retry logic and backoff
- Processes referral batches from CSV/JSON with templated emails
Go Concurrency Goroutines SMTP
Reliability > Features
Observability is not optional
Simple systems fail gracefully — complex ones fail silently
Every performance fix starts with a measurement
A service nobody can debug is a service nobody should ship
- Designing for partial failure in distributed systems
- CDC (Change Data Capture) patterns for event-driven architectures
- eBPF for low-overhead observability in Kubernetes workloads
- Consensus algorithms beyond Raft — Viewstamped Replication, EPaxos