Shiva Shankar Kanike
Research Assistant Ml Systems @University of Maryland Baltimore County
Signup · Get unlimited contacts
WORK HISTORY
Research Assistant Ml Systems @University of Maryland Baltimore County
Washington, DC, US
Built and deployed a scalable satellite imagery ingestion and preprocessing system integrating Landsat 8/9, Sentinel-2, and VIIRS data sources, with automated workflows for metadata validation, tile alignment, and dataset consistency across large geospatial collections.• Built end-to-end ML training and serving workflows on AWS SageMaker with Hugging Face Transformers and LangChain retrieval pipelines, covering experiment tracking, artifact versioning, and automated retraining for geospatial wildfire risk prediction.• Developed Python-based geospatial processing services and batch transformation utilities for large raster datasets, automating wildfire severity metric computation (NBR index), reducing manual preprocessing effort by 80% and cutting fire impact analysis time from days to near real time.• Cut wildfire burn map generation from 14 days to under 1 minute by engineering a pix2pix GAN pipeline on multispectral Landsat and Sentinel-2 imagery, while extending coverage from U.S-only to global satellite datasets.• Researched and applied Vision Transformers (ViT) and diffusion-based image synthesis for high-resolution burn map reconstruction, improving pixel-level segmentation precision by 37% over baseline GAN outputs.
EDUCATION
Mahatma Gandhi Institute of Technology
B.Tech, Electronics and Communication Engineering
University of Maryland Baltimore County
Master's degree, Data Science
ABOUT SHIVA SHANKAR KANIKE
Software Engineer with 2+ years of experience building distributed systems, backend infrastructure, and ML-powered applications. Master\'s in Data Science from UMBC (GPA 3.9/4.0), incoming SWE at Google. I build systems end to end, from REST APIs and microservices to multi-agent orchestration platforms, distributed tracing infrastructure, and real-time collaborative tooling. Recent work includes a 5-agent async system that cut response latency from 15s to under 5s, a trace collector processing 100M+ spans/day, and a RAG pipeline that reduced analyst turnaround from 3 hours to 20 minutes. My stack: Python, C++, Java, TypeScript, FastAPI, Node.js, React, Kafka, Kubernetes, Docker, PostgreSQL, Redis, AWS, GCP, and Azure. Focused on backend systems, distributed systems, ML infrastructure, and full stack engineering.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.