Presentations
Slide decks and talk materials from industry events, covering cloud architecture, microservices, and pragmatic engineering.
Featured Talks
Microservice Scars
A candid look at the real-world challenges and lessons learned from adopting microservices at scale — covering organizational friction, operational overhead, and architectural trade-offs.
DynamoDB in Production
Patterns for running DynamoDB reliably: incremental backups, capacity monitoring, partition-aware design, and cost optimization strategies drawn from production experience.
CloudFormation Deep Dive
Infrastructure as code with AWS CloudFormation — templates, stacks, change sets, and best practices for repeatable, auditable provisioning.
SlideShare Archive
Presentation materials from colleagues and collaborators are available on SlideShare. Topics span microservice architecture, continuous delivery pipelines, and cloud-native design patterns.
- Microservice Scars — Joshua Toth
- Additional presentations by Joshua Toth and Tal Rotbart explore related themes in distributed systems and engineering leadership.
Cloud infrastructure presentations often serve as a bridge between theoretical knowledge and practical application. Engineers who share their experiences at industry events provide invaluable insights that cannot be found in documentation alone. The real value emerges when speakers candidly discuss the challenges they faced, the mistakes they made, and the solutions they eventually discovered. These talks frequently cover architectural patterns, operational strategies, and the nuanced trade-offs that come with building distributed systems at scale. Attendees leave with actionable knowledge that can be applied directly to their own environments, accelerating their learning curve and helping them avoid common pitfalls that others have already navigated.
Running DynamoDB in production requires a thoughtful approach to capacity planning and partition management. Hot partitions can emerge unexpectedly when certain keys receive disproportionate traffic, leading to throttling and degraded performance. Effective monitoring strategies help teams detect these imbalances early, allowing them to adjust read and write capacity before users are impacted. Production patterns that include incremental backups, careful schema design with appropriate sort keys, and secondary global indexes can dramatically improve reliability. These operational practices transform DynamoDB from a simple key-value store into a robust foundation for demanding workloads that must scale gracefully under unpredictable access patterns.
The journey toward microservices architecture often leaves teams with unexpected scars that rarely appear in vendor marketing materials. Organizational friction becomes a significant concern as teams must coordinate across service boundaries while maintaining velocity. Operational overhead multiplies with each new service, requiring sophisticated deployment pipelines, observability tooling, and incident response procedures. The architectural trade-offs between consistency, availability, and partition tolerance become painfully real when systems fail in production. Successful teams learn to embrace these challenges rather than ignore them, building resilience into both their systems and their organizational structures through honest retrospectives and continuous improvement.
Infrastructure as code has fundamentally changed how teams provision and manage cloud environments. Manual configuration processes that were once time-consuming and error-prone have been replaced by repeatable, automated workflows that treat infrastructure as version-controlled artifacts. Tools that define resources in declarative templates allow teams to create, update, and tear down complex environments with confidence. This approach eliminates the drift between environments that plagued traditional operations, ensuring that staging and production remain consistent. The ability to codify infrastructure also enables peer review, change tracking, and rapid recovery from failures, making cloud operations safer and more predictable for teams of any size.