Automatic resolution of Incidents in Service-Now
- Worked on building end-to-end ML model that automatically resolves 30% of incidents raised in ServiceNow, reducing their manual resolution time by over 95%. Technologies: Python, Machine Learning.
- Used GitHub as version control system for code development.
- Performed exploratory data analysis. Technologies: Pandas, Numpy.
- Achieved a 95% accuracy rate in incident prediction by applying Random Forest and Support Vector Machine algorithms.
- Integrated MongoDB database with ML application using Pymongo to manage and store 10,000+ tickets.
- Migrated on-premises ML application to Azure cloud, which resulted in a 20% reduction of infrastructure cost. Technologies used: Docker, Kubernetes, Azure DevOps.
- Deployed Monolithic ML application as separate modules using microservices architecture and containerization.
- Implemented Continuous Integration and Continuous Deployment processes, which increased deployment speed and efficiency by 30%.
- Worked in agile and collaborated with cross-functional teams to research, identify, and analyze solutions.
- Used Jira for agile project management, tracking and prioritizing tasks, and Confluence for documentation.
- Worked on Chatbot using Rasa framework for the users to create, update, view tickets in ServiceNow.
- Used NLP models like Spacy and BERT in development for precise user query understanding.
