GEETHA KONDADASULA
Vijayawada,India
Summary
AI Engineer specializing in GPU-accelerated inference and production AI systems, with hands-on experience building low-latency, high-throughput model-serving platforms using NVIDIA Triton, TensorRT, and Kubernetes. Proven ability to take deep learning models from optimization to reliable, monitored, zero-downtime production deployments. IEEE-published researcher with a strong foundation in computer vision, deep learning architectures, and performance-driven AI infrastructure.
Overview
1
1
year of professional experience
1
1
Certification
Work History
AI Engineer — GPU Inference Platforms
NVIDIA
Santa Clara, USA
09.2024 - 01.2026
- Architected a production-grade GPU inference platform using NVIDIA Triton, enabling low-latency, high-throughput multi-model serving
- Optimized PyTorch / ONNX models with TensorRT, achieving ~35% inference latency reduction while improving GPU utilization
- Designed dynamic batching strategies to balance throughput and latency under strict production SLAs
- Deployed inference services on Kubernetes with GPU scheduling, enabling horizontal scalability and high availability
- Implemented production observability (latency, throughput, GPU utilization) using Prometheus and Grafana
- Supported blue-green and canary deployments, ensuring zero-downtime model updates in live environments
- Tech Stack: NVIDIA Triton, TensorRT, CUDA, PyTorch, ONNX, Docker, Kubernetes, Python, REST, gRPC
Education
Master of Science - Information Studies
Trine University
Detroit, MI08.2024
Bachelor of Technology - Computer Science and Engineering in Artificial Intelligence Specialization
KL University
Vijayawada, IN05.2022
Skills
- Python
- SQL
- NumPy
- Pandas
- PyTorch
- TensorFlow
- Keras
- Scikit-learn
- Convolutional Neural Networks (CNNs)
- LSTM
- YOLOv5
- NVIDIA Triton Inference Server
- TensorRT
- ONNX
- CUDA (usage-level)
- Dynamic Batching
- Docker
- Kubernetes
- MLflow
- Airflow
- CI/CD Pipelines
- Prometheus
- Grafana
- REST
- GRPC
- OpenCV
- AWS (Glue, Bedrock)
- Apache Spark
- PySpark
- Apache Kafka
- Snowflake
Certification
- IBM Data Science Professional Certificate
- Coursera ML with Python Certificate
- Coursera Machine Learning Foundations Certification
Research And Publication
Perception of Age and Gender Detection Using Hierarchical Deep Learning Architecture — IEEE Xplore, Built a CNN-based computer vision system using OpenCV for age and gender classification, Trained and evaluated models on the Adience Dataset, applying transfer learning for improved accuracy, Performed error analysis using confusion matrices to identify limitations and guide model improvements
Timeline
AI Engineer — GPU Inference Platforms
NVIDIA
09.2024 - 01.2026
Master of Science - Information Studies
Trine University
Bachelor of Technology - Computer Science and Engineering in Artificial Intelligence Specialization
KL University