Engineering Blog
Technical writing on computer vision, MLOps, and backend systems.
backendMar 5, 2026
Building High-Throughput Asynchronous Tasks with FastAPI and Redis Event Queues
How to keep FastAPI request handlers thin by offloading heavy AI inference work to Celery workers behind a Redis broker.
mlopsFeb 24, 2026
Optimizing YOLOv8 Object Inference to Sub-10ms Using TensorRT and Docker
Cutting YOLOv8 inference latency down through TensorRT compilation, INT8 calibration, and a container layout that avoids cold-start penalties.
computer visionFeb 10, 2026
Implementing Swin Transformer Blocks from Scratch in PyTorch
A walkthrough of windowed self-attention and the shifted-window mechanism that makes Swin Transformers efficient for dense vision tasks.