Hire Giridhar K. - MLOps Engineer

At Suki, I work on the Ambient Clinical Intelligence product as part of the ML team. My role focuses on building scalable ML systems and pipelines, where I leverage PEFT techniques for efficient model development. I design and develop frameworks for RAG-based and autonomous agents using LangChain and LangGraph, integrating them with ADKs. I also handle model deployment and inference using FastAPI on Serverless instances. On the infrastructure side, I’ve set up cost-efficient environments for model development and testing using managed instances, reducing both training and inference costs through optimized resource utilization and pay-as-you-go strategies. Additionally, I collaborate closely with doctors and clinical teams to build evaluation frameworks using Label Studio. These systems allow clinicians to review model outputs and provide structured feedback, which is then reused for continuous model improvement. I’ve also built on-demand applications using Gradio and Streamlit to help teams visualize model performance and debug issues in real time.

Part of the Infer IoT product, developed a Python-based microservice architecture (“Adapter”) to integrate customer IoT devices with the platform, enabling scalable and reliable data ingestion. Built machine learning models for anomaly detection, device failure prediction, and risk scoring using IoT device data. Lead client engagements, including onsite product demonstrations and technical discussions.

Skills: Back-end Operations · Internet of Things (IoT) · Product Development · Defining Requirements · REST APIs · Docker Products · Back-End Web Development · Git