Samiul (Sam) | kernelshard
Senior Backend Engineer • 6+ Years • Distributed Systems
Important
Precision is freedom. Build systems that tell the truth.
| Category | Technologies |
|---|---|
| Languages | Go · Python |
| Frameworks | gRPC · FastAPI · Django · Gin |
| Data & Streaming | PostgreSQL · Kafka · RabbitMQ · Kinesis · Redis |
| Infra & Ops | Docker · AWS · Prometheus · Grafana |
| Architecture & Patterns | Microservices · Event‑Driven · CQRS · Hexagonal · Saga |
| Area | How I Deliver |
|---|---|
| 🚀 Complexity | Turn ambiguous reqs into decoupled, event-driven microservices with strict API contracts. |
| 📈 Performance | Engineer low-allocation, high-concurrency primitives. Optimize for p99 & throughput before they break. |
| 🔭 Observability | Make structured logs, Prometheus metrics, and SLAs a default—not an afterthought. |
| 🧑🏫 Team Growth | Mentor juniors, enforce strict interface boundaries, and own the on-call rotation. |
Explicit > Magical.
I choose composable, boring libraries over clever abstractions that hide failure modes.
If a service doesn’t expose its internal state via metrics and traces, it is incomplete.
⚙️ forge — Distributed Job Queue Go
- Event-driven architecture: Designed with Kafka messaging to achieve ~40% lower API latency over traditional polling layouts.
- Resilient core: Token-bucket rate limiter, heap-based priority queue, and LRU cache primitives.
- Zero downtime: Worker pools gracefully drain on
SIGTERM— zero lost jobs during deployments.
🌐 expose — Secure Reverse Tunnel Go
- Multiplexed streams: Handles 100s of concurrent tunnels over a single TCP connection.
- Zero-allocation core: Optimized internal proxy loops using
sync.Poolto crush GC pressure during high-throughput transfers.
📦 imagekit-python — Official Python SDK
- Original author & core maintainer serving hundreds of enterprise clients.
- Zero breaking changes across 3 major Python versions — fully thread-safe & async-ready.
Deep in distributed systems — Kafka, gRPC, and the failure modes that come with running services at scale. Also focusing on data consistency via the Saga Pattern.
Open to collabs on distributed systems, infrastructure ownership, or senior backend roles. Let's build something that doesn't fall over.
Built with 🖤, caffeine, and a deep hatred for silent failures.