Staff Applied Scientist working on generative AI and video/content intelligence systems.
I build and ship applied ML systems in production — currently focused on generative video, multimodal content pipelines, and the infrastructure that makes AI-generated media reliable at scale. I write about the technical and economic realities of this space, and I'm increasingly interested in the founder/investor side of where AI-native media and content businesses are headed.
- Generative AI & multimodal systems — applied research and production ML for AI-generated video and content pipelines
- ML infrastructure & data engineering — scalable pipelines (Spark, Airflow, NiFi), model lifecycle management, and the plumbing that makes GenAI systems actually work in production
- The business/economics layer of AI media — how generative video changes cost structures, content supply chains, and creator economics
(Newest first — see Medium for the full archive)
- Optimizing Spark Performance with AQE
- Mastering Data Processing with Apache Spark's Catalyst Optimization
- 10 Best Practices for Building Scalable Data Pipelines
- Local-LMOps-Monitoring — monitoring tooling for locally-hosted LLM operations
- multi-document-summarizer — multi-document summarization as a service
- RAGs — RAG implemented from scratch
- object-detection-with-transformers — DETR-based end-to-end object detection
- Languages & ML: Python · PyTorch · TensorFlow · Scikit-learn
- GenAI/NLP: Agentic AI · LLMs · RAG · Langchain · Langgraph · Multimodal pipelines
- Data Engineering: PySpark · Airflow · NiFi · Kafka
- Infra: Docker · FastAPI · Opensearch · AWS (EMR, EC2, S3, SageMaker, Glue)
Open to conversations on generative video, AI content systems, and where this space is headed. Book time via Topmate.