I’m Rohit Chivukula, a Machine Learning Engineer at NextGen Federal Systems.
I build production GenAI and ML systems from research and modeling to deployment and MLOps on AWS. I also have a background in 3D medical imaging and deep learning weather models!
- Architecting GenAI systems with AWS Bedrock, vector search (OpenSearch), and GraphRAG on Neo4j for secure document understanding.
- Building AI agents and MCP tools that orchestrate retrieval, analysis, and workflows inside chat-style interfaces.
- Owning ML infra with AWS CDK, SageMaker, and containerized services for repeatable, observable deployments.
- GenAI Document Systems: RAG/GraphRAG pipelines over large technical corpora using Bedrock, OpenSearch, and Neo4j.
- Deep Learning Weather Forecasting: Aurora weather model inference on ERA5/GRIB data for global forecasts.
- 3D Medical Imaging (WVU): 3D U-Net models on 1TB+ PET/CT data for lesion segmentation and related diagnostic tasks.
- Multimodal Federated Learning in Healthcare: Co-authored a survey covering 50+ works in FL + medical AI. Find the title Multimodal Federated Learning in Healthcare: a Review on arXiv.