Thiveya D

• Kubernetes:

Deployed trained models into production systems with tools like Docker containers or Kubernetes clusters.

Implemented containerization technologies such as Docker and Kubernetes for faster delivery cycles.

Used K3S Kubernetes service along with Azure Arc to perform High Availability for ML models throughout the clusters and served scalability based on the CPU percentage of the devices.

Used Linkerd as service mesh for observability, monitoring and security purposes and traffic routing.

Used edge devices like Intel NUC, Nvidia Jetson to perform high availability and scalability.

• CT Device Software – Pepsico Client:

It is on integrating Sodastream Drinkstation Devices with IoT Central for PepsiCo, enabling seamless transmission of telemetry data, consumable tracking, and execution of Firmware OTA and OS updates and also implemented security protocols using CA certificates.

Technologies Used: Python, IoT Central, Fast API

• Device Observability:

It is on optimizing the aggregation and analysis of metrics, logs, and traces from IoT Edge modules to a cloud-based dashboard for enhanced visualization and monitoring, leveraging Open Telemetry standards.

Technologies Used: Python and Machine Learning techniques, alongside Open Telemetry SDK, to drive efficient data processing and insights generation.

• MTK - Master Tool Kit:

It is an end-to-end solution for data preparation, model training and deploying models to an edge device so it can be used across Accenture irrespective of any dependencies

Some of use cases I've worked on are,

ML Use Case: Development of an advanced object detection model, leveraging YOLOv5 and OpenCV, to accurately identify quality discrepancies in tablet blisters through real-time video analysis. Utilized Python and sophisticated image augmentation techniques to enhance machine learning algorithms, ensuring high precision in automated quality control processes.

Medical Synthetic Data Generator to generate synthetic data to build a model with GAN network.

Azure IoT Defender : Designed and deployed an IoT Edge Device and integrated Defender Micro agent to ensure comprehensive log collection, in alignment with Azure Log Analytics Workspace. Crafted automated playbooks within Azure Sentinel to efficiently convey incident information related to Security alerts.

Technologies Used: Python, Bash, Azure IoT, Azure Sentinel, and ARM Templates

• Done Infrastructure Automation projects, boosting efficiency, accelerating delivery times, and heightening client contentment.
• Developed and implemented PowerShell and Python scripts to enhance build, deployment, and maintenance processes.
• Innovated by converting manual operations into automated systems, refining project workflows with scripting and sophisticated tools.
• Automated ticket handling from BMC, streamlining the update process, saving time, and improving customer satisfaction using Python Selenium, Tkinter, pandas, and win32com.
• Pioneered an Automated Ticket Creation system with Selenium, enabling entry-level staff to manage tasks effortlessly, reducing dependencies, and conserving resources.
• Appreciated and Awarded for Consistent Support and Effort

• Engineered an intuitive employee performance portal featuring dynamic graphs and tables, enhancing user comprehension through sophisticated drill-down capabilities. Leveraged Django, Python, HTML, CSS, JS, Highcharts, Plotly, and MongoDB.
• Underwent Trainings on: Python, PowerShell, Web Technologies (HTML, CSS, Bootstrap, PHP, JavaScript, MySQL ) , Orchestration Tools(vRO, MS-Scorch), ITSM Tools (ServiceNow,BMC)