Open Source Summit + Embedded Linux Conference North America 2026

As a Linux Kernel Developer at Ampere Computing, I focus on optimizing the scheduler for high-density ARM64 systems. My work culminates in a patch merged for the Linux 7.0 release that refines avg_idle tracking a critical metric the scheduler uses to decide how long to search for an idle CPU before giving up. In my session "The Hidden Cost of Sleep", I will break down the try_to_wake_up() path to show how even minor inaccuracies in idle-time accounting lead to poor CPU selection and increased cache misses. I’ll explain how my Linux 7.0 optimizations [commit
36ae1c45b2cede] specifically reduce the 'search cost' during wakeups, directly improving the responsiveness of AI inference workloads. By sharing raw performance data and trace analysis, I’ll demonstrate why getting the wakeup path right is the only way to achieve the deterministic performance needed for autonomous AI agents and scalable trust infrastructure.

LLM tokens are expensive. With context windows expanding to 200K+ tokens, a single API call can cost several dollars & in production systems handling thousands of requests, these costs compound quickly.
Most optimization efforts focus on model selection or prompt engineering, but the context itself often contains massive redundancy.

Headroom is an open-source Python library (https://github.com/chopratejas/headroom) that sits between your application and your LLM provider, transparently optimizing context before it reaches the model.
The core insight is simple: LLM contexts—especially in agentic workflows—are filled with repetitive tool outputs, verbose JSON arrays, and boilerplate that consumes tokens without adding proportional value

Headroom introduces novel concepts such as reversible compression, cache aligners, compression routers, and even persistent memory

Real-world results:
- 50-90% token reduction on typical agentic workloads
- Drop-in integrations for LangChain, OpenAI, Anthropic, and any OpenAI-compatible provider
- Zero code changes required when using the proxy server