Principal ML/AI Architect • LLM/RAG/Agentic AI • MLOps/LLMOps • Data/Cloud Engineering • Research Scientist

📍 Stony Brook, NY

I build production-grade AI systems—especially agentic workflows, RAG with evaluation + guardrails, and LLMOps/MLOps on AWS / Azure / GCP—turning research prototypes into monitored, secure, cost-aware services.

• Agentic AI systems (LangGraph, CrewAI, AutoGen) and tool-graph orchestration
• RAG in production: retrieval design, reranking, evals (RAGAS/TruLens), guardrails, and governance
• LLMOps/MLOps: model serving, observability, drift/eval jobs, and reliability SLOs
• Data platforms & streaming: event-driven pipelines, lakehouse patterns, and scalable feature delivery
• AI infra: Kubernetes + Terraform + GPU scheduling/autoscaling
• Distributed systems foundations: fault tolerance, consistency trade-offs, RPC, and scalable service design

AI / Agentic: LangGraph • CrewAI • AutoGen • LlamaIndex • Semantic Kernel • n8n
RAG / Retrieval: RAGAS • TruLens • Guardrails • Pinecone/Weaviate/Qdrant/Milvus/Chroma/pgvector • rerankers • ColBERT-style retrieval
MLOps / LLMOps: MLflow • Ray Serve • BentoML • KServe • Seldon • OpenTelemetry • Prometheus/Grafana
Cloud GenAI: AWS Bedrock • SageMaker JumpStart • Azure AI Studio/Prompt Flow/Azure OpenAI • GCP Vertex AI Agent Builder
Data / Governance: Databricks/Delta • Iceberg/Hudi • dbt • Great Expectations • Evidently AI • CDC/ELT
Infra: EKS/AKS/GKE • Helm • Terraform • Docker • Vault/KMS
Distributed Systems: Raft/Paxos/PBFT • sharding/replication • leader election • 2PC • gRPC/Protobuf • backpressure & load balancing concepts
OS / Systems: Linux • processes/threads • concurrency primitives • memory & I/O fundamentals • TCP/IP basics • profiling/debugging

Medical Imaging: 3D Slicer • DICOM • NIfTI • ITK/VTK • segmentation & annotation workflows
GIS / Remote Sensing: QGIS • ArcGIS • Google Earth Engine • GDAL/OGR • Rasterio • GeoPandas • Shapely • Fiona • Sentinel-2 • Landsat 8/9 • NDVI/EVI/SAVI • Land Surface Temperature • DEM Processing • AHP Multi-Criteria Analysis
Languages: Python • SQL • PySpark • TypeScript/JavaScript

• Consensus & replication: Raft, Paxos / Multi-Paxos, PBFT; leader election; failure handling
• Transactions & coordination: sharding, replication strategies, 2PC concepts for cross-partition operations
• RPC & networking: gRPC + Protobuf, service boundaries, timeouts/retries, load balancing fundamentals
• Concurrency: threads vs async I/O, synchronization basics, race conditions/deadlocks (and avoidance patterns)
• OS fundamentals: Linux, process lifecycle, scheduling concepts, memory/I/O basics, performance profiling

• Reproducible PyTorch experiments packaged with MLflow + cloud storage
• RAG over lab literature/annotations/notes to accelerate hypothesis generation
• Data-quality + GPU-usage guidelines for shared research infrastructure

• Azure Event Hub → Databricks (PySpark) → Delta → Azure ML pipeline for high-volume 5G events
• MLflow + Azure DevOps + Kubernetes for monitored rollouts and retraining

• Real-time ICU analytics: video/vitals → Kafka/PySpark → AWS (EC2/Lambda/SageMaker/Redshift/DynamoDB)
• Containerized CV/multimodal models with Docker + EKS for clinician dashboards

• Debian-CFI-Census — Contributor
• MIMIC-III Benchmarks (clinical ML) — Contributor
• Semantic Search for Medical Practitioners (Cohere) — Contributor
• Text Summarization via OpenAI/Cohere — Contributor
• YOLOv4-OpenCV-CUDA-DNN — Contributor
• Vector search prototypes (Qdrant/Chroma) + Docker cluster trials — Contributor
• 3D Slicer Digital Assets for AI Research — Contributor

Check my pinned repos for the most up-to-date work.

• Urban Heat Island Analysis & Mitigation Mapping (Urban Planning)
  Land Surface Temperature extraction from Landsat 8/9 thermal imagery • UHI zone classification using standard deviation thresholds • LULC correlation analysis • Weighted Mitigation Priority Index combining thermal anomaly, population density, and green space deficit
  5 source modules, unit tests, data download utility

• Multi-Criteria Flood Risk Assessment (Environmental/Climate)
  DEM hydrological preprocessing (sink filling, flow direction, flow accumulation, TWI) • AHP-weighted flood susceptibility mapping (7 criteria with consistency validation) • Socioeconomic vulnerability assessment (population, infrastructure, SVI) • Composite risk index with scenario analysis for multiple return periods (10yr–500yr)
  6 source modules, comprehensive AHP unit tests

• Crop Health Monitoring with Sentinel-2 NDVI (Agriculture/Remote Sensing)
  Multi-temporal Sentinel-2 data loading with cloud masking and 20m→10m resampling • 7 vegetation indices (NDVI, EVI, SAVI, NDRE, NDWI, GNDVI, CIre) • Double-logistic phenology curve fitting • Anomaly detection (z-score, Isolation Forest, LOF) for crop stress identification • Field-level zonal statistics and change reporting
  6 source modules, extensive unit tests

• Brain Tumor Segmentation & 3D Volumetric Analysis (Neuro-oncology / AI Training Data)
  End-to-end DICOM → NIfTI ingestion pipeline in 3D Slicer • Semi-automatic brain tumor segmentation using Segment Editor (threshold painting, grow-from-seeds, margin expansion) on multi-modal MRI (T1, T1-Gd, T2, FLAIR) • 3D volumetric rendering and tumor volume quantification • Label-map export for deep learning model training (nnU-Net / MONAI compatible) • Custom 3D Slicer Python scripted module for batch segmentation QA and inter-rater consistency checks
  Hands-on with 3D Slicer Segment Editor, Markups, Volume Rendering, and scripted module development

• CT Lung Nodule Annotation & AI-Ready Dataset Pipeline (Pulmonology / AI Research)
  DICOM loading and lung windowing in 3D Slicer • Semi-automatic pulmonary nodule segmentation using threshold-based region growing and manual refinement • Annotation of nodule characteristics (size, shape, margin, texture) via custom Markups and JSON metadata export • Batch conversion of segmentations to NIfTI label maps with standardized naming for AI model ingestion • Integration with a Python post-processing script for COCO-style dataset formatting and train/val/test splitting
  Hands-on with 3D Slicer DICOM module, Segment Editor, Markups, Python interop, and dataset curation for AI pipelines

Publications

• Facial Emotions and Behaviour Detection using DNN (Scopus, 2021)
• Text Recognition from Images (Tesseract + gTTS) (IJARCS, 2020)
• Dinuc2mer: Deep Learning Approach for Coding Region Prediction (BioCHEM, 2024)
• Adaptive RAG Orchestration and Tool-Graph Optimization for Enterprise LLMs (2025)
• Adaptive Importance Sampling for Rare Congestion in a Two-Stage Queue (2025)
• Using ML to Identify Dinucleotide Motif Periodicity in Genomes (2025)

Patents

• Monitoring Facial Emotions and Behavior Using DNN (AU 2021102414, 2021)
• Provisional: Adaptive RAG Orchestration and Tool-Graph Optimization (2025)
• Provisional: Multimodal ICU Deterioration Prediction via Streaming Pipelines (2023)

• M.S. Computer Science (Thesis) — Stony Brook University (2023–2024)
  Coursework: ML, CV, NLP, Big Data, DB Systems, LLMs, RAG & Agentic AI, Distributed Systems (Raft/Paxos/PBFT, gRPC, Blockchain)
• B.Tech CSE (Thesis) — REVA University (2017–2021), GPA 9.78

• President, Graduate Student Organization — SBU (May 2024 – Jun 2025)
• Chief Steward, CWA Local 1101 GSEU (Mar 2024 – Present)
• Board of Directors, ASA — SBU (May 2024 – May 2025)

• Email: bindubhagavareddy@gmail.com
• LinkedIn: bindu-bhargava-reddy-chintam
• GitHub: @bhargavchintam
• Phone: +1 (934) 221-7847