- Published on
Building real-time intent recognition achieving sub-500ms latency with 70% accuracy improvement, and implementing Query AutoComplete using Ghosting approach with N-gram models and fine-tuned GPT-2.
Senior ML Engineer
Building production ML systems at scale. 5+ years delivering NLP, Computer Vision, and MLOps solutions.
NLP & LLMs
Intent Detection, Transformers, GPT Fine-tuning
MLOps
SageMaker, MLFlow, Kubernetes, CI/CD
Computer Vision
YOLOv8, SAM, Object Detection
Edge AI
TensorRT, Quantization, Model Optimization
Latest Posts
Deep dives into ML systems, architecture decisions, and lessons learned.
- Published on
How I designed and deployed an automated MLOps system that reduced ML cycle time by 90%, managing end-to-end model lifecycles including training, versioning, serving, and drift detection.- Published on
Reduced weekly construction site reports time by 70% using ML models (YOLOv8, UNet, SAM, Transformers) for materials detection. Built event-driven integration handling 1000+ concurrent jobs.- Published on
System design exploration on detecting harmful content using NLP, Computer Vision, and ML at scale. Covers sentiment analysis, hate speech detection, content filtering, and real-time moderation strategies.- Published on
Improving edge inference speed by 25-30% via quantization (QAT, TFLite, TensorRT, OpenVINO) and pruning. Building DeepOps for real-time edge device management reducing model management time by 80%.
Subscribe to the newsletter