Salma Shaik
Research Software Engineering @ Microsoft AI
Bengaluru, India
Actions
Salma is currently an Applied Researcher at Microsoft Research — Systems Innovation and a freelance Machine Learning Engineer, who has previously worked at 5+ AI-driven startups helping them deploy and scale production-ready machine learning systems. She loves diving deep into machine learning and multi-agent systems, with a focus on algorithm design, large-scale LLM workloads, and robust, generalizable solutions. Her best days are when she builds dev tools that boost productivity, and improve cloud software reliability, collaborating with organizations like MIT CSAIL, CMU ML, Microsoft, and IISc!
Links
Area of Expertise
Topics
Building scalable cloud systems with AI and DDD
In this session, Salma Shaik, Applied Researcher at Microsoft Research, will present how multi-agent large language models (LLMs) can revolutionize incident management in DevOps. Drawing from her experience in automating incident triage for large-scale cloud services at Microsoft, she will delve into the implementation of AI-powered workflows that reduce Mean Time to Mitigation (MTTM) by 35%. Attendees will gain insights into how semantic routing, adaptive SLAs, and context-aware prioritization can optimize incident resolution processes, providing real-world examples of AI-driven automation within DevOps pipelines. This session will cover the integration of AI into existing DevOps practices, enhancing cloud reliability and performance through intelligent incident triage.
Enriching source code with contexual data for local AI autocompletion
We'll visit steps for designing and training Python-aware code autocompletion models from scratch using Transformer-based architectures; talk about inference optimization while ensuring high throughput, and discuss strategies for model quantization.
Next-gen Security Automation: Leveraging Microsoft Defender XDR and Sentinel for Threat Mitigation
In this session, Salma will explore how Microsoft Defender XDR provides a comprehensive solution for safeguarding workplace environments through real-time threat detection, advanced security automation, and seamless incident response. Drawing from her experience with cloud systems and multi-agent frameworks, she will demonstrate how integrating Defender XDR with existing workplace infrastructures can elevate security standards and enhance operational efficiency.
The session will dive deep into the use of Microsoft Defender for Endpoint, Defender for Office 365, and Defender for Identity, showing how to automate security management and response across endpoints and identities. Salma will also cover the integration of Microsoft Sentinel for advanced analytics, centralized threat monitoring, and efficient incident triage, ensuring that security teams can react quickly to evolving threats.
Attendees will gain insights into the real-world application of Microsoft Defender XDR and how to configure it for a holistic security approach that anticipates, detects, and mitigates threats autonomously.
Building resilient DevOps pipelines with AI for scalable cloud reliability
In this session, we will explore how DevOps practices are enhanced by multi-agent LLM systems to automate and optimize incident triage in cloud environments. Drawing from my work at Microsoft Research, I’ll share how I implemented a multi-agent framework that reduced Mean Time to Mitigation (MTTM) by 35% in large-scale cloud services. By leveraging semantic routing, adaptive SLAs, and context-aware prioritization, we transformed traditional manual workflows into dynamic, AI-powered systems. This session will focus on the intersection of DevOps, incident management, and machine learning, showing how AI-driven automation can streamline cloud reliability operations, enhance system performance, and minimize downtime. Attendees will gain practical insights on how to integrate AI-based solutions into their DevOps pipelines to ensure faster incident resolution and improve overall system scalability and resilience.
Building Autonomous AI Systems for Scalable Cloud Operations
In this session, Salma will dive into the cutting-edge realm of autonomous AI systems and how they are revolutionizing cloud operations. Focusing on the integration of AI agents and multi-agent systems, she will discuss their application in automating complex cloud workflows, including incident management, fault localization, and real-time mitigation.
The session will explore the use of Large Language Models (LLMs) and reinforcement learning techniques for self-healing cloud environments. Through practical examples, Salma will demonstrate how AI agents can autonomously make decisions, adapt to dynamic cloud conditions, and optimize system reliability with minimal human intervention.
Attendees will gain a deeper understanding of the key challenges involved in building scalable, robust autonomous systems for cloud operations, as well as the tools and frameworks that enable seamless orchestration between AI agents and cloud services.
AI-driven automation in cloud incident management
As cloud environments scale, manual incident triage and resolution become bottlenecks. This session explores how AI-powered automation can transform incident management workflows in cloud-native systems. Using ML models and cloud infrastructure like Azure, we’ll discuss how AI can automate classification, prioritize incidents, and route tasks in real-time, reducing mean time to resolution (MTTR) and ensuring service reliability.
We’ll cover the integration of AI models for incident triage, dynamic SLA management, and real-time prioritization, enabling self-healing systems that scale with cloud demands. Key topics include:
• Automated incident classification and escalation with LLMs.
• AI-driven dynamic routing based on severity levels and historical incident data.
• Predictive scaling to handle high volumes of incidents using cloud-native tools (e.g., Azure, Kubernetes).
• Data privacy and security in AI-driven cloud automation.
AI-driven automation in cloud environments
Learn how to integrate AI-powered automation into cloud infrastructure to improve scalability and efficiency. This session will explore how ML models can be used to automate critical processes in cloud environments, reducing manual overhead and improving system reliability. We’ll focus on designing resilient systems capable of handling large-scale data and dynamic workflows.
Agentic AI for DataOps
In this session, Salma Shaik will explore how agentic AI is transforming DataOps by automating and optimizing data workflows within Microsoft Fabric. The session will focus on how autonomous agents, powered by reinforcement learning and AI-driven orchestration, can streamline complex data engineering tasks and enhance the scalability of cloud data operations.
Attendees will gain insights into how agentic AI can autonomously orchestrate DataOps tasks across Microsoft Fabric, reducing manual interventions and improving overall data processing efficiency. The session will highlight real-world use cases and best practices for integrating AI agents into cloud data engineering workflows.
Salma Shaik
Research Software Engineering @ Microsoft AI
Bengaluru, India
Links
Actions
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top