Specializing in ML Systems, Distributed Platforms, and High-Performance Infrastructure
I sit at the intersection of System Engineering and Applied Machine Learning. I optimize hardware utilization for large models, manage compute resources in heterogeneous environments, and build distributed ingestion pipelines processing multi-terabyte datasets. My engineering goal is to construct highly reliable and efficient Big Model-as-a-Service (BMaaS) platforms.
|
High-performance inference benchmark and observability tool for NVIDIA Triton Server. Monitors P50/P99 latency and uses NVML for active VRAM & hardware telemetry.
|
Autonomous Generative AI agent for stranded travelers. Implements state-persistent routing graph workflows and hybrid vector search filters using MongoDB Atlas.
|
|
Distributed data platform processing 5TB+ daily with 100+ concurrent jobs. Orchestrated with Apache Airflow and Apache Spark on Kubernetes pods.
|
Event-driven microservice system implementing distributed idempotency, optimistic locking, and message queue patterns to ensure strict transactional consistency.
|
|
Real-time data ingestion and analytics pipeline. Streams stock market telemetry, runs validation checks, and persists logs into Cassandra databases.
|
Predictive ML model cataloging and analyzing historical viewer statistics. Optimizes content trend ratings using advanced feature engineering pipelines.
|
| Languages | Python, C++, Go, Java, SQL, Bash |
| AI & Systems ML | PyTorch, NVIDIA Triton, NVML, LangGraph, LangChain, Vector Search |
| Cloud & Ops | Kubernetes, Docker, AWS (SQS, DynamoDB), Redis |
| Data & Streaming | Apache Spark, Apache Kafka, Apache Airflow, MongoDB, Cassandra |
🎀 Fun Fact: I paint, drink a lot of coffee, and write high-performance code—sometimes all in one night! 💻✨