Building High-Performance Distributed Systems in Modern C++
Real-World Patterns with Boost.Asio & Beast
The power of Modern C++ for designing low-latency, scalable distributed systems is significant; however, most of the real-world development experience is not found in typical textbooks and/or industry-wide conferences. As such, the purpose of this open-source educational session is to provide a walk-through of the design/implementation of production-level distributed services with the use of modern C++, specifically utilizing both Boost.Asio and Boost.Beast.
This session will be based upon the architectural patterns employed by large scale systems that produce millions of events per second while maintaining a very low latency/availability constraint. The topics will include:
- Asynchronous I/O Design
- Structured Concurrency
- Back-Pressure Handling
- Request Pipelining
- Connection Lifecycle Management
- Failure Detection
- Retries
- Graceful Degradation
In addition, this session will cover some of the most common pitfalls associated with developing production level distributed services written in C++. These pitfalls include:
- Accidental Blocking
- Hidden Contention
- Memory-Lifetime Bugs
We will also explore ways to avoid these common pitfalls when designing your distributed system.
Unlike most other sessions at conferences which introduce yet another framework (which typically requires a lot of overhead), this workshop will focus on providing attendees with composable building blocks and design techniques which attendees can directly apply to their own distributed service designs (RPC systems, Control Planes, Gateways, Edge Deployed Services). As such, the material presented in this workshop is geared towards seasoned C++ developers wishing to move beyond syntax and learn about how to design high-performance, resilient distributed systems.
Samaresh Kumar Singh
Samaresh Kumar Singh is a Principal Engineer at HP Inc. with more than 21 years of experience architecting and building complex, high-performance, and highly available distributed systems. Throughout his career, his primary focus has been the design and implementation of large-scale, low-latency systems in modern C++ (C++14/17/20/23), leveraging Boost, Boost.Asio, and Boost.Beast to develop resilient networking, concurrency, and service orchestration frameworks.
Samaresh has led the development of high-throughput, fault-tolerant microservices and distributed control planes that operate across cloud, hybrid, and edge environments. His work emphasizes correctness under concurrency, efficient resource utilization, robust failure handling, and horizontal scalability—often in systems processing millions of events per second with strict latency and availability requirements. He has extensive experience designing custom RPC layers, asynchronous I/O pipelines, distributed caching, leaderless and consensus-based coordination, and service discovery mechanisms in production-grade systems.
In recent years, Samaresh’s work has expanded to edge-native AI and intelligent systems, where he applies his distributed-systems background to build scalable platforms for Edge AI/ML, agentic AI systems, and GenAI/LLM-based services. His current focus includes explainable and trustworthy AI at the edge, federated and privacy-preserving learning, secure hybrid cloud-edge orchestration, and efficient deployment of AI workloads on resource-constrained devices.