Deepak Raj
@deepakraj0103
GEN AI ENGINEER at KRATO SOLUTIONS
India
Deepak Raj is a Generative AI Engineer with over 1.3 years of experience delivering production-grade LLM systems, RAG pipelines, and multimodal AI platforms. He possesses expertise in retrieval optimization, LoRA/QLoRA fine-tuning, and scalable cloud deployment. His professional background includes roles as a Data Scientist and Quality Engineer, where he successfully implemented vision pipelines and anomaly detection systems to improve operational efficiency and accuracy.
Experience
GEN AI ENGINEER
KRATO SOLUTIONS
Architected RAG pipelines processing 10,000+ HTA/NICE PDFs, shrinking execution time 60% using Azure OpenAI, Bedrock, Databricks. Optimized retrieval via chunking + reranking, elevating response accuracy 68% to 91% and lowering hallucinations 35%. Established RAGAS evaluation and implemented prompt validation with fallback logic.
DATA SCIENTIST
AISPRY
Led YOLOv8 + DeepLabv3+ vision pipelines on 50,000+ images, reaching 93% segmentation precision. Designed DenseNet201/ResNet50 IVF models, increasing prediction accuracy 28%. Delivered Flask + React microservices enabling <500ms inference latency. Streamlined MLflow workflows.
QUALITY ENGINEER
SBC PVT LTD
Instituted SPC + anomaly detection systems, lowering defect rate 15% and saving $120K annually. Crafted analytics dashboards eliminating 50% manual reporting effort. Conducted root-cause modeling across 3 facilities, reducing process variability 12%.
Education
Saveetha Engineering College, Chennai
BACHELOR OF ENGINEERING
MECHANICAL ENGINEERING
Licenses & Certifications
Professional Course in Data Science & AI
360DigiTMG
IT-ITES Sector Skills Council NASSCOM Certification
NASSCOM
Claude Code on AWS Bedrock – Anthropic / AWS Generative AI Certification
Anthropic / AWS