I'm Lahiru Niroshan Sathsara. I administered production Oracle databases for core banking, and I build data projects around reliability, clear pipelines and useful analytics. I'm looking for data engineering and data analytics roles.
Right now:
Oracle DBA → Data engineer
Building reliable data systems
Oracle
Python
PostgreSQL
Kafka
Docker
DuckDB
Streamlit
Pandas
Power BI
Spark
SQL
GitStreaming, reconciliation, controls and analytics. The data is simulated; the engineering and measurements are real.
An OpenF1-fed streaming architecture stores race data in PostgreSQL and processes Kafka events with Spark into Cassandra. Airflow orchestration is in progress; dashboards are planned.
OpenF1 · Kafka · PostgreSQL · Spark · Cassandra · Airflow
Reconciles a simulated core-banking database against its reporting replica. Cheap partition checks run first, and row hashing only runs where drift is suspected.
Python · PostgreSQL · Docker · Streamlit
A control tower for billing exceptions, discount decisions, SLA monitoring and customer risk, backed by a DuckDB warehouse.
Python · DuckDB · SQL · Streamlit
Synthetic marketplace data used to explore orders, pricing, restaurant performance, driver operations and churn risk.
Python · SQLite · Scikit-learn · Streamlit
Grouped by the job they do. What I'm still learning is kept separate on purpose.
Where the data lives
OraclePostgreSQLSQLDuckDBHow it gets where it is going
PythonPandasKafkaHow people see it
DockerStreamlitPower BIGitOn the roadmap
AirflowSpark
Cassandra
dbtFour things I can take on from day one, each backed by a project you can open.
Ingest, store and process data with Kafka, Spark, PostgreSQL and Docker, built to run end to end.
See: F1 Real-Time PipelineFind silent drift between systems, classify it, and leave evidence someone can act on.
See: DataGuardTurn tables into decisions with Streamlit, Power BI and DuckDB, with metrics defined once.
See: Billing Assurance, Uber EatsOracle 19c/24ai on production servers: backup and restore with RMAN, query tuning, monitoring scripts.
Experience: VS Information SystemsWorking on live banking databases taught me to trust data only after checking it.
Where the data comes from, who depends on it, and what a bad day looks like.
Checks at each hop, so problems are caught where they start.
Containerised, repeatable, and easy to rerun after a failure.
Clear numbers, stated limits, and notes the next person can follow.
I started in IT support, then administered production Oracle systems supporting core banking. That shaped how I build: check the data, understand what can fail, and make the results useful. My next step is data engineering, with analytics built on top.
I'm looking for data engineering, analytics and database roles. Send a message, or take my CV with you.