Headline

AI Engineer @ dAIgnostiX | ML Engineer | GenAI • RAG • LLM • MLOps | Python • FastAPI • Docker | M.Tech (IISc Bangalore, ’23)

About

AI Engineer (Team Lead) with 3 years of experience building production-grade GenAI and ML systems. I specialize in designing and deploying LLM-powered RAG pipelines including hybrid retrieval architectures with vector databases (Qdrant) and LLM orchestration that solve real-world problems in production environments. At dAIgnostiX, I built an end-to-end RAG system that personalizes AI-generated radiology reports to match individual writing styles, reducing manual editing time by 40%, alongside designing evaluation frameworks to measure model reliability in clinical settings. Beyond GenAI, I've shipped high-uptime AI products using FastAPI, Docker, and MongoDB, architected event-driven systems with real-time drift detection and feature stores, and applied deep learning to diagnostic imaging (segmentation, detection) in healthcare. I currently lead a team of 4 AI engineers, working hands-on across model development, system design, and code reviews. Recently, I built an end-to-end real-time cryptocurrency price prediction system covering feature engineering, drift detection, model registry, and CI/CD, deployed as 9 microservices on a Kafka-compatible event-driven architecture. I'm always interested in connecting with others working on GenAI, applied ML, and production AI systems.

Location

Hyderabad

AI stack

codex

opencode

Work experience

AI Engineer (Team Lead) at dAIgnostiX, Jul 2023 - Present · 3 yrs 2 mos, Bengaluru, Karnataka, India, • Built an end-to-end RAG pipeline personalizing AI-generated radiology reports to match radiologists’ writing styles, using hybrid retrieval with Qdrant and OpenAI’s models, reducing manual editing time by 40%. • Designed a reference-anchored evaluation framework to measure style-adaptation quality in isolation from radiologists’ independent clinical edits, improving reliability of automated report generation. • Designed an AI model orchestration service using FastAPI, RabbitMQ, MongoDB for distributed inference across services, bringing multiple AI products to production. • Led design and deployment of an event-driven subscription service with flexible plans, expanding monetization by 70%. • Lead a team of 4 AI engineers, overseeing model development and code reviews. • Developed and deployed segmentation models for aortic dissection and aneurysm detection, achieving IoU 0.83 and 0.92 with 94% sensitivity and 89% specificity, reducing reporting time by nearly 60%. • Built pneumonia detection and YOLO-based localization models with mAP 0.7, 90% sensitivity and 88% specificity, cutting evaluation time by approximately 30%.

Education

MTech, Signal Processing, Indian Institute of Science (IISc)

BTech, Electrical Engineering, Bhilai Institute of Technology (BIT), Durg