I'm a Backend & Data Engineer based in the United States, specializing in high-performance REST APIs, asynchronous microservices, and large-scale ETL/ELT data pipelines. I design systems that handle millions of events per day — from ingestion through transformation to analytical delivery — with a focus on reliability, observability, and clean architecture.
- 🔧 Building production APIs with FastAPI and Django REST Framework
- 🔄 Designing and operating ETL / ELT pipelines at scale (batch & streaming)
- ☁️ Cloud-native infrastructure on AWS and GCP
- 📊 Enabling data teams with warehouse-first architectures using dbt, Airflow, and Spark
- 🧪 Advocate for test-driven development, typed Python, and CI/CD automation
A production-grade FastAPI service with async request handling, JWT auth, Redis caching, Celery background workers, and PostgreSQL persistence — deployed behind CloudFront and AWS ALB.
Schema-per-tenant PostgreSQL isolation, RBAC middleware, pluggable installed apps split across Core / Business / API layers, Celery task queues, and external service integrations.
End-to-end batch and streaming pipeline: 6 source types → Kafka/Airbyte ingestion → Airflow DAG orchestration → Spark + dbt transformation → Great Expectations quality → Snowflake warehouse → BI / reverse-ETL serving.
Sub-second real-time pipeline processing 500k+ events/minute: producers → Kafka topics with Schema Registry → Flink jobs (enrich, dedup, aggregate, anomaly detect) with RocksDB state → ClickHouse OLAP + Redis counters → Grafana dashboards.
| Category | Technologies |
|---|---|
| Primary | Python 3.12, SQL (PostgreSQL dialect, BigQuery SQL, Snowflake SQL) |
| Secondary | Bash / Shell, Go (basics), TypeScript (basics) |
| Python Ecosystem | asyncio, typing, pydantic, dataclasses, mypy, black, ruff |
| Framework | Use Case | Key Libraries |
|---|---|---|
| FastAPI | Async REST APIs, microservices, ML serving endpoints | Uvicorn, Gunicorn, SQLAlchemy async, Alembic, Pydantic v2, python-jose, passlib, slowapi |
| Django | Full-stack SaaS apps, admin-heavy platforms, multi-tenant systems | DRF, drf-spectacular, SimpleJWT, Celery, django-tenants, django-storages |
| Flask | Lightweight internal tools, simple webhook receivers | Flask-RESTful, Flask-SQLAlchemy |
| Layer | Technologies |
|---|---|
| Orchestration | Apache Airflow 2.x, Prefect, Dagster |
| Batch Processing | Apache Spark (PySpark), dbt Core, Pandas, Polars |
| Stream Processing | Apache Flink, Kafka Streams, Spark Structured Streaming |
| Message Bus | Apache Kafka, AWS SQS/SNS, Redis Streams |
| CDC / Replication | Debezium, AWS DMS, Airbyte |
| Transformation | dbt (staging → intermediate → marts), custom PySpark jobs |
| Data Quality | Great Expectations, dbt Tests, Soda Core |
| Type | Technologies |
|---|---|
| Relational | PostgreSQL (primary), MySQL, SQLite (testing) |
| OLAP / Warehouse | Snowflake, Google BigQuery, ClickHouse, AWS Redshift |
| Lakehouse | Delta Lake, Apache Iceberg, AWS Glue Data Catalog |
| Cache / KV | Redis, Memcached |
| Search | Elasticsearch, OpenSearch |
| Document | MongoDB, DynamoDB |
| Object Store | AWS S3, GCS, MinIO |
| Category | Technologies |
|---|---|
| AWS | EC2, ECS, EKS, RDS/Aurora, S3, Glue, EMR, Lambda, SQS, SNS, CloudWatch, IAM |
| GCP | BigQuery, Cloud Composer (Airflow), Dataflow, Cloud Run, GCS |
| Containers | Docker, Docker Compose, Kubernetes (EKS/GKE), Helm |
| IaC | Terraform, AWS CDK, CloudFormation |
| CI/CD | GitHub Actions, GitLab CI, ArgoCD, Jenkins |
| Tool | Purpose |
|---|---|
| Prometheus | Metrics collection and alerting rules |
| Grafana | Dashboards — infrastructure, pipeline health, API latency |
| Loki | Log aggregation and querying |
| Sentry | Error tracking and performance monitoring |
| OpenTelemetry | Distributed tracing across microservices |
| DataDog | APM and full-stack observability (enterprise projects) |
| Airflow UI / Flower | Pipeline and task monitoring |
| Concern | Approach / Tool |
|---|---|
| API Style | REST (primary), GraphQL (selective), WebSocket (real-time) |
| Documentation | OpenAPI / Swagger (drf-spectacular, FastAPI native) |
| Authentication | JWT (SimpleJWT, python-jose), OAuth2, API Keys |
| Rate Limiting | Redis-backed token bucket, slowapi (FastAPI) |
| Serialization | Pydantic v2, DRF Serializers, marshmallow |
| Testing | pytest, pytest-asyncio, factory_boy, Hypothesis, httpx |
Kafka → Flink → ClickHouse pipeline processing 500k+ events/minute with real-time anomaly detection and sub-second dashboard latency. Built deduplication using bloom filters and stateful windowing with RocksDB-backed Flink state.
Stack: Python · Apache Kafka · Apache Flink · ClickHouse · Redis · Grafana
Airflow-orchestrated ETL platform ingesting from 12 source systems into Snowflake, serving a 15-person analytics team. Introduced dbt for SQL transformation governance and Great Expectations for automated data quality SLAs.
Stack: Python · Apache Airflow · dbt · Snowflake · Great Expectations · AWS S3 · Airbyte
Async REST API backend for a B2B SaaS product — 100k+ daily active users. Implemented async SQLAlchemy with connection pooling, Redis cache-aside pattern, and Celery for background report generation. P99 latency under 120ms.
Stack: FastAPI · PostgreSQL · Redis · Celery · AWS ECS · Terraform · GitHub Actions
PostgreSQL schema-per-tenant Django application supporting 80+ enterprise tenants. Designed custom TenantMiddleware for schema routing, role-based permission system, and Stripe billing integration with webhook handling.
Stack: Django · DRF · PostgreSQL · Celery · Stripe · Docker · AWS RDS