Venkatesh Maligireddy

• Developed AI-driven Incident Management Platform: Optimizing incident management at PayPal through AI-powered automation and prediction – Auto Summary, Communications (with 90% accuracy), etc.
• Building Centralized Data Lake for SRE: Collaborating on a centralized data repository for SRE platforms, enabling intelligent data analysis and decision-making leveraging combined insights.
• Designed and deployed a workflow-based Developer support automation leveraging a combination of ChatOps and Agentic AI solutions. This enabled automatic resolution of 50% of developer queries without human intervention.
• Designed and completed a POC of an Agentic AI platform internal to the Organization that supports State Persistence and Human-in-the-loop based on LangGraph and CrewAI frameworks.

• Led the design and implementation of a scalable, first-of-its-kind Incident Management Copilot leveraging open-source and third-party LLMs for data analysis, Q&A, and live journaling, with a focus on DLP (Data Loss Prevention) and legal compliance.
• Designed and Engineered a comprehensive solution to utilize incident data from diverse sources such as ticketing systems, Slack conversations, and triage tools, enhancing the functionality of Incident Management Copilot. Employed a mix of methodologies including real-time webhooks, batch processing, and caching to cater to various use cases.
• One of the first in the company to experiment and showcase the advantages of applying Gen-AI technology on Incident data leveraging its reasoning and comprehensive abilities using OpenSource LLM Models for efficient Incident Response.
• Engineered and deployed a cutting-edge ML-driven Incident Prediction system, achieving early detection for 45% of P0/P1 incidents within 6 weeks, showcasing end-to-end engineering excellence in solution development, from discovery to frontend and backend implementation with streamlined data fetchers.
• Developed, Launched, and operationalized an automated AI-enabled issue intake system in Slack, streamlining the process for over 4.5k users across PayPal, and collaborating across teams on requirements, design, and integration.
• Established and currently leads the Tech Council, having conducted 9 sessions to date and extending its reach to the entire APAC community.
• One of the core members that developed the Chaos Engineering Platform at PayPal. Enhanced it by extending fault injection to non-standard container VMs and orchestrated 4 Gamedays in Production with the Infra team to enhance AVI load balancer resiliency.
• Developed a novel Enterprise Resiliency score framework using chaos experiments, including industry research, design, and front-end/back-end implementation, ensuring scalability for future score definition updates.

• Led cross-team collaboration to develop an end-to-end novel solution addressing uncorrelated FCIs, resulting in the elimination of over 30,000 Tier 1/2 Failed Customer Interactions (FCIs) per day and attributing approximately 16 million FCIs annually to the correct teams. This also resulted in significant reduction of customer calls, enhancing productivity of customer support agents.
• Designed and engineered "Sitevitals", a comprehensive visualization tool to monitor the health of various domains within the company's stack. Implemented upgrades to monitoring and alerting pipelines, resulting in a significant improvement with a 92% onboarding success rate.
• Introduced a new metric - "Time to Triage" for incident response efficiency evaluation into the incident ticketing flow.
• Led ChatOps platform for automated incident management in Slack, including an innovative interactive messaging feature and optimized workflows. Implemented a secure external endpoint for internal application (first of its kind), gaining architecture committee approval.
• Championed the adoption of a new, scalable application within the technology stack, enabling a "bot-first" approach for SRE products, streamlining incident management.

• Full-Stack Development Expertise: Successfully led the development of the Regional Parity Dashboard (RPD) from conception to launch, demonstrating proficiency in both frontend and backend technologies. Utilized skills in coding, debugging, and problem-solving to ensure the functionality and performance of the application across various components.
• Effective Cross-Team Collaboration: Facilitated seamless collaboration among a diverse team, working across different time zones and engaging with multiple stakeholders. Demonstrated strong communication and coordination skills to integrate feedback, align objectives, and deliver a comprehensive solution that addresses the company's disaster recovery readiness goals.
• Led discovery sessions for the Active/Active initiative, collaborating with diverse leaders to shape product ideation and strategy.
• Leveraged NLP/ML expertise while mentoring interns on a chatbot project, fostering leadership through on-the-spot decision-making based on project knowledge, individual strengths, and future goals.

• Designed and implemented new features (bug notifications, change approvals, status checks) for an internal ChatOps tool, enhancing user experience and gaining exposure to various systems and processes.
• Developed a Search feature prototype comparing SQL vs. NoSQL technologies, gaining insights into design patterns and technology selection criteria.

Non-blind Deblurring: Handling Kernel Uncertainty with CNNs

• Played a key role in designing and developing custom deep convolutional neural network models in Torch language to address kernel uncertainty, crucial for non-blind image deblurring in the project.
• Independently developed and implemented all coding requirements for model training and inference.

• Worked as an intern in the Site Reliability Engineering(SRE) - Tools Team
• Designed and built two webpages of an Internal Tool using AngularJs, which went live during my Internship period

• Learned image processing algorithms on Monocular Visual SLAM for Autonomous Vehicles.
• Evaluated existing algorithms in the field and suggested a new efficient algorithm that could improve the performance of the existing algorithm