I'm a Data Engineer who builds the infrastructure that AI systems run on — pipelines, lakehouses, and data platforms designed to be consumed by both dashboards and models.
My work sits at the intersection of data engineering and applied AI: I care as much about how clean, reliable data gets to a model as I do about the model itself.
- 🏆 Google GenAI Exchange 2025 — Top 100
- 🔍 Open to ML Data Engineering internships & full-time roles
- 🛠️ Currently building lakehouses, RAG pipelines, and event-driven data systems
Also working with:
Apache Spark·PySpark·Airflow·dbt·Databricks·Delta Lake·BigQuery·Snowflake·Polars·Pandas·LangChain·RAG·Vector Embeddings
Python Apache Kafka PySpark AWS S3 Snowflake dbt Airflow Docker
Built a Python producer that streamed 6M+ clickstream events into Apache Kafka (KRaft), simulating live user traffic with rate limiting and offset tracking.
A PySpark Structured Streaming job cleaned and transformed the events before writing them as partitioned Parquet files to an S3 data lake. From there, the data moved into Snowflake via External Stages, where dbt built marts for conversion funnels and cart abandonment.
Airflow orchestrated the complete pipeline, with the entire stack running in Docker.
Python Celery Redis Transformers Flower
Engineered a distributed batch-processing pipeline using Celery and Redis as both message broker and cache.
Interview questions were cached and persisted via Redis, reducing response latency by 40%. Transformer-based NLP models powered automated AI interview sessions, while Flower provided real-time monitoring of the distributed pipeline.
Python GCP Redis RAG Docker
Built a full recommendation pipeline using vector embeddings, a vector database, and RAG to power product discovery.
Deployed inference services on GCP Cloud Run, with Pub/Sub handling event-driven messaging and Redis providing low-latency caching.
The entire platform was containerised and hosted on GCP.
