Projects

Synchromesh Profile Scanner

Developed an automated defect detection and measurement system using machine learning and computer vision techniques to identify defects in mechanical components during different stages of testing.

The system processes component images and detects structural defects while accurately measuring defect dimensions using image processing and deep learning methods.

Impact:
• Reduced manual inspection effort by approximately 93% in repetitive testing workflows.
• Achieved around 90% defect detection accuracy.
• Enabled faster defect analysis and improved testing efficiency.

Key Contributions:
• Designed the image processing and machine learning pipeline.
• Implemented automated defect detection and measurement algorithms.
• Deployed the solution as an event-triggered pipeline with partial CI and full CD implementation for a robust and automated workflow.

Tech Stack:
Python, OpenCV, TensorFlow, Image Processing, CI/CD Pipelines, Databricks

User Logs Analysis and Dashboarding

Developed a data analysis and visualization system to monitor application usage, detect anomalies, and support root cause analysis using user log data.

Impact:
• Improved issue diagnosis and root cause identification for system failures.
• Increased user satisfaction by more than 50% through faster issue resolution.
• Enabled monitoring of application performance and usage trends.

Key Contributions:
• Processed and analyzed large volumes of user log data.
• Built dashboards to visualize application usage patterns and anomaly indicators.
• Provided actionable insights to guide system optimization and performance monitoring.

Tech Stack:
Python, SQL, Databricks, Data Visualization

Vehicle Telematics Analytics

Designed an automated analytics pipeline to process raw vehicle telemetry data and generate structured reports for engineering analysis.

Impact:
• Reduced manual analysis time and repetitive work by approximately 70%.
• Improved data interpretation and visualization for engineering teams.
• Enabled faster insights from large telemetry datasets.

Key Contributions:
• Built a pipeline to ingest raw telemetry files and transform them into structured datasets.
• Implemented automated report generation for key metrics and analysis results.
• Added logging mechanisms to track pipeline performance and error handling.

Tech Stack:
Python, SQL, Databricks, Data Processing Pipelines

Parts Imaging Automation and Report Generation

Developed an automated system for processing component images and generating inspection reports using machine learning and computer vision techniques.

Impact:
• Automated repetitive inspection workflows for component imaging tasks.
• Improved consistency and speed of report generation.
• Reduced manual effort involved in image analysis and documentation.

Key Contributions:
• Built image processing workflows for analyzing component images.
• Integrated automated report generation for inspection outputs.
• Designed the pipeline to support scalable processing of image datasets.

Tech Stack:
Python, OpenCV, Deep Learning, Image Processing, Automated Reporting