Ramanathan Parameshwaran
Principal Forward Deplyed Engineer (AI) at Amazon Web Services (AWS)
Karnataka, India
Summary
🤩
Rockstar🎓
Top SchoolRamanathan Parameshwaran is a Principal Forward Deployed Engineer (AI) with 11 years of experience building production-grade ML systems across retail, education, and finance. He blends deep learning research with pragmatic engineering to deliver low-latency, high-throughput solutions—from multimodal recommendation and retail shelf monitoring to real-time meeting intelligence. Ramanathan has strong MLOps chops, contributing notable integrations for Weights & Biases and Hugging Face Transformers to improve experiment logging and model artifact workflows, and has worked on YOLO-based instance segmentation with TensorRT acceleration. Comfortable in cross-functional product teams, he moves models from prototype to scale and optimizes pipelines for real-world constraints. Based in Karnataka, India, he pairs a mechanical engineering foundation with hands-on AI productization experience that surfaces in both open-source contributions and production deployments.
10 years of coding experience
11 years of employment as a software developer
Bachelor of Technology (B.Tech.) Mechanical Engineering, Bachelor of Technology (B.Tech.) Mechanical Engineering at Shanmugha Arts, Science, Technology and Research AcademyMachine Learning Nanodegree Machine Learining, Machine Learning Nanodegree Machine Learining at Udacity
English, Telugu, Hindi, Kannada, Tamil
Github Skills (36)
transformers10
pytorch10
integrate10
python10
weight10
machine-learning10
detectors10
integrations10
yolov810
detector10
weighting10
mlops10
instance-segmentation10
deep-learning10
wandb10
Programming languages (3)
Github contributions (5)
lucasjinreal/yolov7_d2
Jul 2022 - Aug 2022
🔥🔥🔥🔥 (Earlier YOLOv7 not official one) YOLO with Transformers and Instance Segmentation, with TensorRT acceleration! 🔥🔥🔥
Role in this project:
ML Engineer
ML EngineerContributions:20 commits, 1 PR, 3 comments in 1 month
Contributions summary:Ramanathan primarily focused on integrating a Weights & Biases (W&B) logger into the YOLOv7 deep learning model's demo script. Their commits added a custom W&B formatter and inference logger to visualize instance segmentation results, including bounding boxes, masks, and confidence scores, within the W&B platform. The user also modified the demo script to support logging, added options for W&B project and entity configuration, and corrected label mappings for compatibility with W&B visualizations.
accelerationtensorflowtransformersobject-detectioncomputer-vision
wandb/wandb
Oct 2022 - Jun 2024
The AI developer platform. Use Weights & Biases to train and fine-tune models, and manage models from experimentation to production.
Role in this project:ML Engineer
ML EngineerContributions:50 reviews, 17 PRs, 64 pushes in 1 year 8 months
Contributions summary:Ramanathan primarily contributed to the integration of the `ultralytics` library for YOLOv8 within the Weights & Biases platform. This included developing a `WandbCallback` class, adding various callbacks for logging metrics, model artifacts, and images during the training process. The user's contributions focused on enabling seamless tracking and visualization of YOLOv8 model training runs within the W&B ecosystem, encompassing metric logging and model artifact management. Further contributions included updates to support new releases of the Langchain library and the autologging of OpenAI and Cohere model usage and metrics.
pythoncollaborationtensorflowhyperparameter-tuningcli
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.