← Back to Portfolio

Technical Blog

Deep dives into building production AI/ML systems

•10 min read

Building Brckt: Real-Time Tennis Analytics with RAG

How I built a 5-stage RAG pipeline for real-time sports analytics, reducing LLM latency from 3s to 400ms using streaming and quantization

#RAG#LLM#Real-time#Production#FastAPI
•12 min read

Building Cascade: An Intelligent LLM Router That Cut API Costs by 60%

How I built a production-ready ML system to optimize LLM routing using DistilBERT, semantic caching, and real-time cost tracking

#LLM#Machine Learning#Cost Optimization#Production#MLOps