Parag Gupta

The Problem: Modern systems generate tons of telemetry data — metrics (CPU, memory), logs (errors, events), and traces (request flows). Traditional tools look at these separately, which leads to noisy, siloed alerts and "alert fatigue," making it hard to spot the real issue.

The Solution: This patent introduces an AI-powered system that fuses all these signals into a unified structure, builds a cause-and-effect graph of events, and then works backward to pinpoint the true root cause of an incident (instead of just the noisy symptoms). What makes it stand out is a learning loop -- every time an SRE validates the system's prediction, it learns and adapts, becoming smarter with each incident.

The Innovation: In essence, this creates an AI Ops brain that doesn't just correlate data, but truly understands causation, and improves continuously like a human expert.

A highly-available Kubernetes cluster setup with multiple master nodes to ensure zero downtime for critical applications. This project includes:

• Redundant control plane with 3+ master nodes using etcd clustering
• HAProxy load balancing for API server traffic
• Custom node auto-scaling for handling varying workloads
• Advanced network policy configuration for pod-to-pod communication
• Custom monitoring stack with Prometheus, Grafana, and Alertmanager
• Security hardening with pod security policies and network isolation

The solution provides 99.99% uptime for mission-critical services while optimizing resource usage during peak and off-peak hours.

A comprehensive set of Terraform modules for deploying and managing infrastructure across multiple cloud providers. Key features include:

• Reusable modules for AWS, Azure, and GCP resources
• Consistent naming, tagging, and security policies across all deployments
• Automated testing of infrastructure using Terratest
• CI/CD pipeline integration for infrastructure deployments
• State management and locking with remote backends
• Drift detection and automated remediation

This approach reduced infrastructure deployment time by 70% and improved compliance with security policies across all environments.

A comprehensive monitoring and observability solution for microservices architecture. The platform provides:

• End-to-end visibility across the entire application stack
• Custom exporters for application-specific metrics
• Intelligent alerting with PagerDuty integration and alert routing
• Automated dashboard generation for new services
• Historical data analysis for capacity planning
• Service level objective (SLO) monitoring and reporting

This solution reduced mean time to detect (MTTD) by 60% and mean time to resolve (MTTR) by 45% for production incidents.

An AI-powered application for analyzing virtual machine metrics and providing actionable insights. Features include:

• Automated data collection from Cortex metrics database
• Machine learning models for anomaly detection in VM performance
• Interactive visualizations of resource usage patterns
• Natural language processing for querying metrics data
• Prediction of resource needs based on historical patterns
• Recommendation engine for VM right-sizing

This tool identified optimization opportunities that resulted in a 30% reduction in cloud infrastructure costs.

A platform that converts natural language instructions into executable Ansible playbooks. Key capabilities include:

• Natural language processing to understand automation requirements
• Automatic generation of Ansible playbooks from text descriptions
• Interactive editing and validation of generated playbooks
• Integration with existing Ansible inventory and variables
• Execution tracking and reporting
• Knowledge base of common automation patterns

This tool enabled non-technical teams to implement infrastructure automation, reducing the time to automate common tasks by 80%.

A comprehensive solution for managing and optimizing cloud infrastructure costs. Key components include:

• Automated resource scheduling for non-production environments
• Right-sizing recommendations based on utilization patterns
• Reserved instance and savings plan analysis
• Multi-account and multi-cloud cost visualization
• Budget alerting and forecasting
• Tagging compliance enforcement

This initiative reduced cloud spending by approximately $120,000 annually while maintaining or improving application performance.