Agus Richard Lubis
Senior Software Engineer · Backend & Data
+6285710276393 · agus.richard21@gmail.com · linkedin.com/in/agusrichard
agusrichard.medium.com · github.com/agusrichard · agusrichard.com
Professional Summary
Senior Software Engineer (Backend & Data) with 7+ years of experience building distributed systems and high-throughput data pipelines in production. Led ETL re-architectures that tripled ingestion capacity and cut end-to-end processing from days to minutes — unlocking new client onboarding and the revenue growth that reliable infrastructure makes possible. Takes full ownership from defining requirements and managing backlogs through delivery, testing, and client onboarding — sustaining continuous integration and deployment cycles that adapt fluidly to evolving client requests. Led a team of 6 engineers, contributed to hiring, and stepped in when delivery was at risk — bringing that same sense of ownership to AI adoption: using it deliberately to compress learning cycles and amplify productivity, never to replace engineering judgment or abdicate the design, the tradeoffs, and the craft. Active technical writer with 25+ articles on Medium — writing is how understanding becomes durable.
Skills & Proficiencies
Python · Go · Rust · JavaScript · TypeScript · Scala · System Design · Software Architecture · gRPC · Microservices · ETL / ELT · AWS · Kubernetes · Docker · Terraform · GitHub Actions · PostgreSQL · DynamoDB · Redis · Redshift · Snowflake · BigQuery · MongoDB · Elasticsearch · SQL · Airflow · Kafka · Spark · dbt · Pandas / NumPy · FastAPI · Django · Flask · Celery · Prometheus · Grafana · RabbitMQ · SQS · SNS
Work Experience
- Served as technical project lead for a team of 6 engineers — owning system design, technical specifications, code reviews, and client communication; participated in the engineering hiring loop to define role requirements and evaluate candidates.
- Took sole ownership of a stalled, overdue project — defining technical requirements, managing the backlog, and conducting code reviews to drive it to a successful production delivery.
- Decomposed a monolithic ETL into three independent DAGs, enabling concurrent multi-client ingestion and reducing end-to-end processing from days to minutes while tripling ingestion capacity.
- Replaced legacy ingestion with async API calls and dynamic task generation, eliminating out-of-memory failures and reducing failed DAG runs from 2–3 per week to zero; introduced idempotent retry mechanisms with active alerting.
- Led migration of existing clients to the new ETL platform and defined a standardized onboarding procedure adopted across all property management integrations, ensuring seamless data ingestion from day one.
- Redesigned the Login and Registration flows in C# and ASP.NET Core Web API, improving API structure, usability, and visual consistency.
- Maintained product stability through targeted bug fixes throughout the contract, resolving issues with minimal disruption to end users.
- Decomposed a monolithic application into 3–5 independent microservices across an 8–10 engineer team, decoupling deployments and enabling each service to be scaled and updated without system-wide risk.
- Implemented SOAP-based integrations with external systems and replaced point-to-point calls with event-bus-coordinated workflows, automating bidirectional data exchange and guaranteeing cross-service data consistency.
- Leveraged Go Goroutines to stream live position updates for ~100 vessels onto a real-time Google Maps view without blocking the main thread, giving operators immediate fleet-wide situational awareness.
- Built a structured logging layer auditing all cross-service data exchanges, reducing incident investigation time and providing a verifiable data integrity trail across service boundaries.
- Delivered end-to-end Login and Registration functionality across both Web and Mobile platforms, maintaining a consistent user experience.
- Built web scraping pipelines for 2–3 clients, automating extraction of hundreds of records per run and replacing manual data collection; profiled production applications to identify and resolve performance bottlenecks.
- Mentored a group of university students on software architecture and engineering best practices, guiding design decisions and improving code quality.
Featured Projects
ICU Vitals Stream Go · Rust · Kafka · PySpark · TimescaleDB · Delta Lake · MinIO | Built an end-to-end real-time streaming pipeline: Go simulator publishes vital signs to Kafka; a Rust consumer (Tokio, dashmap, sqlx) computes NEWS2 scores per patient with sub-millisecond latency and persists to TimescaleDB for live Grafana dashboards. Ran PySpark Structured Streaming jobs alongside the Rust consumer to write 5-min and 1-hour aggregates to Delta Lake on MinIO, serving a Flask-backed ward KPIs dashboard — two complementary stores for two distinct access patterns. |
Real Estate Data Pipeline Python · Airflow · AWS MWAA · AWS Lambda · Polars · Snowflake · Terraform | Designed and deployed a production ELT pipeline on AWS: MWAA orchestrates Lambda functions that ingest from a 2.2M-row Kaggle CSV and the RentCast live API, transform with Polars, stage as Parquet on S3, then MERGE into a Snowflake star schema. Provisioned the full AWS stack (MWAA, Lambda, S3, IAM, CloudWatch) with Terraform and wired CI validation via GitHub Actions — zero manual infrastructure steps, reproducible from a single `terraform apply`. |
Python & Golang gRPC Todo Python · Go · gRPC · Protobuf · Microservices | Built a gRPC-based todo service with both a Python and a Go server implementation, sharing a single Protobuf schema — demonstrating cross-language RPC design and schema-first API contracts. Companion project to the "Build a Microservice App Using gRPC, Python, and Golang" article series on Medium; architecture decisions are documented inline to bridge theory and working code. |
Machine Learning from Scratch Python · NumPy · Jupyter Notebook · Machine Learning | Implemented core ML algorithms from first principles — linear regression (gradient descent and normal equations) and K-nearest neighbors — without high-level frameworks, building deep intuition for the underlying mathematics. Structured as an installable Python package with a pytest suite, mirroring production library conventions and verifying algorithmic correctness through automated tests. |
Volunteer Work
- Mentored junior developers through code reviews and architecture guidance — helping them make sound design decisions and ship production-quality features.
- Authored and maintained end-to-end test suites, establishing a QA baseline for student-built applications.
- Contributed across two independent engagement cycles (2021 and 2023), demonstrating sustained commitment to developer education and raising the quality bar across multiple student cohorts.
Education
Certifications
Publications
Full list: agusrichard.medium.com