Herlangga Yusuf Syailendra

Backend / Data / Infrastructure Engineer

Building scalable backend systems, data pipelines, and cloud infrastructure with a relentless focus on performance, reliability, and cost efficiency.

Reduced latency from ~4s to <100ms via backend and infra optimization.

Lowered cloud costs by ~40% through architecture & query improvements.

Built observability stacks achieving 5x faster anomaly detection.

Tech Stack

Backend

Python Rust Go Node.js

Data

ClickHouse BigQuery Athena

Infrastructure & Cloud

AWS GCP Docker Linux NGINX

Observability

Grafana

Systems & Architecture

High-Volume Email Pipeline

AWS SES → SNS → Service → ClickHouse
  • Processed 500K+ events (delivery, bounce, engagement).
  • Handled 50K+ emails with strong deliverability guarantees (SPF, DKIM, DMARC).
  • Designed an idempotent, retry-safe ingestion architecture.
Email Pipeline Architecture

Monitoring Email from SES Dumped to Clickhouse for low-cost lead generation and engagement tracking.

Data Platform Modernization

  • Migrated from Athena to BigQuery to drastically improve query performance and cost control.
  • Built robust OLAP pipelines for internal analytics.
  • Engineered OLAP → OLTP synchronization to power real-time dashboards and microservices.

Comprehensive Observability Stack

Cloud Run → Log Router → BigQuery → Grafana → Discord
  • Centralized logging and monitoring for highly distributed services.
  • Implemented query-driven metrics to detect anomalies swiftly.
Latency Monitoring

Website Latency and Load Monitoring

Token Usage Dashboard

Analyzing Token Usage & Cost for AWS Bedrock Wrapper

Other Notable Systems

  • AWS Bedrock Wrapper: Built an OpenAI-compatible API Gateway for AI services (OpenAI / Ollama → Bedrock) using Rust, ClickHouse, and SQLite. [View Repository]
  • Distributed Rendering: Configured multi-node distributed Blender rendering infrastructure.
  • n8n Automation: Built complex email processing workflows integrating Listmonk and Discord.
n8n Automation Flow

n8n Triggered from Email → Listmonk → Discord → Review → Send