Systems Architect & Founder | Low-Latency Backend Systems, Analytical Storage Engines & Distributed Computing.
- High-Performance Storage & Systems: Designing analytical database engines, column-store platforms (Parquet/Arrow), and asynchronous I/O architectures (
io_uring,tokio). - Constrained & Edge Machine Learning: CPU-bound local inference pipelines, quantized sub-1B architectures, and strict JSON schema generation (
llama.cpp,ONNX Runtime). - Photonic & Alternate Compute: Exploratory work in non-electronic computing mechanics, optical perceptrons, and physical state switches.
- Cloud & Distributed Infrastructure: Microservice topologies, event-driven streaming (NATS), and cost-optimized compute runtimes (AWS Graviton, container orchestration).
- Analytical Database Engine Components: Custom Rust storage abstractions built for high-throughput, low-latency analytical queries.
- Optical Perceptron: Exploratory photonic computation architectures implementing angle-tunable weights for non-electronic state switches.
- Constrained Extraction Pipelines: High-speed token-level grammar validation and structural parsing engines.
- GitHub: github.com/surajbhan

