Keeping the Pager Human: Balancing AI Self‑Healing Dreams with Operational Duty

A late‑night outage flooded hundreds of thousands of users with 500 errors. Redis logs showed closed sockets, VPC Flow Logs filled with new connections, and CloudWatch warned that the NAT Gateway was at full capacity. An AI chat assistant pointed to a recent third‑party failover change, uncovering a small code bug that triggered a connection storm. Each AI insight handed the investigation back to a human, and every human judgment primed the AI for the next clue in an ongoing feedback loop.

Using this real incident use-case as our guide, we will explore today's AI‑powered SRE tools: alert clustering, anomaly detection, automated runbooks, self‑healing platforms and more. You will see where AI accelerates triage and where an engineer's eye remains essential. Attendees will learn practical techniques to turn raw telemetry into clues, rebuild incident timelines, confirm root causes, and craft RCAs their teams can trust. We will conclude by mapping the next horizon for AI in SRE, outlining the skills and guardrails that keep humans firmly and strategically in the loop.

Disaster Recovery Is the Kingdom of Uncertainty

Your disaster recovery plan isn’t going to work! At least not the way you expect. Recent months showed how “perfect” preparations can collapse when a service provider incident hits and reality drifts off the playbook. Sometimes the plan is robust, but the switch that should activate it can’t be reached (did someone say AWS console login page?).

This talk is about the most underrated DR capability: improvisation. I’ll share hands-on, real-world stories of responding to third-party outages in real time - moments where only deep, end-to-end system understanding, plus a steady head under pressure, made recovery possible. You’ll hear about the night I wired my personal laptop into production to stop revenue from leaking, and the time a temporary client-side SSL change helped bypass a CDN outage, plus a DNS cache making everything worse.

You’ll leave with practical tactics for the next outage, and the mindset that chaos doesn’t read your documentation.

Detective Engineering Case Study: AI-Driven Clue Hunting from Side Effects to Root Cause Evidence

We will examine the thought process and workflow I followed while investigating an outage that impacted hundreds of thousands of users. Beneath the flood of 500-errors was concealed a deeper problem: Redis logs reported closed sockets, VPC Flow Logs displayed thousands of connections, and CloudWatch showed a saturated NAT Gateway. The breakthrough came from an AI‑chat insight about a recent third‑party failover‑strategy change, which pointed directly to a small code bug that triggered a massive connection storm.

In this session you will learn how to treat metrics and logs as evidence, filter out noise, and adopt a detective’s mindset to uncover the true root cause. You will leave with practical techniques to reconstruct incident timelines, leverage AI for hidden insights, apply investigative methods to complex infrastructure failures, and, ultimately, produce an effective RCA.

Accelerating Load Times: A Materialized View and Server-side Composition Case Study

How did I improve the loading of our application by hundreds of percent? I will talk about our architectural change implementation details, and will teach how we leveraged practical design patterns using Redis, AWS Cloudfront, microservices and observability tools to get top-level performance.

This talk is going to be based on a series of 3 articles that I recently published in our engineering blog.

https://medium.com/@yedidya.s/a0f44ca0c55f?source=friends_link&sk=df6acf57392227451224863330261724

Exploring the Depths: Unraveling a Production Incident Caused by a Node.js Memory Leak

In this captivating lecture, I will take you on a journey through a real-life incident involving a performance issue in a new microservice. Drawing from personal experience, I will share how Google Chrome became my trusty companion in debugging the microservice's memory and ultimately uncovering the root cause of the problem.

During the session, I will provide a concise and easily comprehensible overview of memory leaks, using fascinating Node.js examples to illustrate key concepts. By the end of this 30-minute talk, every attendee will gain a solid understanding of why memory leaks occur and acquire practical skills to efficiently debug them using the powerful Google Chrome devTools.

Building Resilience: A Journey of Crafting and Validating Our Disaster Recovery Plan

In the ever-evolving landscape of business continuity, organizations strive to ensure minimal data loss and swift recovery in the face of disruptive events. In this talk, I embark on a compelling journey, exploring the intricate process of crafting and validating a DRP. Through a real-world use case I led at OwnID, I will delve into the practical application of RPO, RTO, Terraform modules, Github actions, execution plans and various DRP architectures, shedding light on the critical decisions, challenges, and successes that arise along the way.

Accelerating Authentication Architecture: Taking Passwordless to the Next Level

How did I improve the authentication process for millions of users by hundreds of percent? I will talk about our architectural change implementation details, and will teach how we leveraged practical design patterns using Redis, AWS Cloudfront, microservices and observability tools to get top-level loading time performance of our passwordless-based authentication solution.

During this talk, I will describe a change that had a significant impact on how customers perceive us as their authentication mechanism. Serving major enterprise companies and well-known brands, we must strive for cutting-edge performance. I will discuss practical architectural changes, design patterns, and various cutting-edge technologies, such as Passkeys, WebAuthn and FIDO2. The true value of this talk lies in integrating all these components, allowing you to explore how familiar tools can be combined with theoretical design patterns to create a remarkable architecture.

Turning Chaos into Code: My IaC Transformation Experience

In this talk I will take you on a journey through the process of transforming our company's manual infrastructure management into a streamlined, automated marvel using Terraform. Witness the power of Terraformer as I guide you through the intricate process of reverse engineering our entire AWS infrastructure and translating it into Infrastructure as Code (IaC) with Terraform. Learn from my experience, explore the challenges I overcame, and understand the tangible benefits of this monumental shift towards infrastructure automation. If you're seeking insights into IaC adoption and the tools that can make it happen, this talk is a must-attend event.

Automate to Elevate: Debugging Your Local JS Code Seamlessly on Customer’s Production

This session unveils a streamlined approach to debug a local version of javascript code on customer production websites, employing the Chrome API response override feature and rollup.js custom plugins. I’ll outline a three-step solution, automating the process of replacing the minified javascript code in a readable and debuggable javascript file on-the-fly.

By integrating seamlessly with your code build process, developers can make and test code changes locally, reflecting them instantly on relevant production websites. This efficient workflow not only elevates the debugging experience but also provides valuable insights into your code performance on real websites, showcasing the power of Chrome API response override and rollup.js in enhancing frontend development cycles.

This session is based on my following article:
https://medium.com/ownid-engineering/automate-to-elevate-debugging-your-local-js-code-seamlessly-on-customers-production-dd26972c06cb?source=friends_link&sk=2a68811e3d33fb4cb5a2e4f7939f44f5

Mastering Public APIs: Lessons from Real-World Implementation and Best Practices

Embark on a journey through the lifecycle of a thriving public API, leveraging a compelling case study from our startup. This talk delivers actionable insights, best practices, and strategies to conquer common challenges in public API development.

Whether you're a seasoned developer or new to public APIs, our case study — exposing a public API for an existing internal API without refactoring, duplicating, or risking the entire project, with a short deadline— offers a fresh perspective.
Join me for a succinct exploration of lessons learned and the mastery of the ever-evolving realm of public API implementation.