Blog | Mathiya (Mathisha Angirasa)

eBPF & Kernel-Bypass Networking in Distributed Cloud Architectures - Cover image for article on Systems Architecture

Systems Architecture📅 30-DAY EPOCH • 30d Left

Oct 1, 2026 9 min read

eBPF & Kernel-Bypass Networking in Distributed Cloud Architectures

A deep architectural dissection of bypassing standard Linux TCP stack bottlenecks using eBPF and eXpress Data Path (XDP) for sub-microsecond cloud networking.

eBPFLinux KernelXDP

Mathisha

READ DEEP DIVE

Autonomous Multi-Agent Swarms & Model Context Protocol (MCP) - Cover image for article on AI & Multi-Agent

AI & Multi-Agent⏳ EPHEMERAL • 13d Left

Sep 30, 2026 7 min read

Autonomous Multi-Agent Swarms & Model Context Protocol (MCP)

An in-depth guide to building resilient, decentralized multi-agent topologies communicating over standardized Model Context Protocols (MCP) with deterministic tool sandboxes.

Agentic AIMCP ProtocolLLM Swarms

Mathisha

READ DEEP DIVE

High-Performance WebGPU Compute Shaders for In-Browser Neural Inference - Cover image for article on WebGL & Graphics

WebGL & Graphics⏳ EPHEMERAL • 12d Left

Sep 29, 2026 8 min read

High-Performance WebGPU Compute Shaders for In-Browser Neural Inference

An authoritative architectural breakdown on executing INT4/FP16 quantized transformer weights at 60+ tokens/sec directly inside modern browser client environments using WGSL compute pipelines.

WebGPUCompute ShadersWGSL

Mathisha

READ DEEP DIVE

Speculative Decoding & KV-Cache Compaction in Multi-Tenant LLM Serving - Cover image for article on AI & Multi-Agent

AI & Multi-Agent⏳ EPHEMERAL • 11d Left

Sep 28, 2026 7 min read

Speculative Decoding & KV-Cache Compaction in Multi-Tenant LLM Serving

An architectural deep-dive into speculative draft routing and virtual memory KV-cache compaction to shatter the memory-bandwidth bottleneck in production LLM inference.

LLM ServingSpeculative DecodingvLLM

Mathisha

READ DEEP DIVE

Zero-Cost Abstractions & Memory Safety in Ultra-Low Latency Rust Engines - Cover image for article on Rust & Performance

Rust & Performance📅 30-DAY EPOCH • 26d Left

Sep 27, 2026 6 min read

Zero-Cost Abstractions & Memory Safety in Ultra-Low Latency Rust Engines

How Rust’s ownership model, compile-time borrow checker, and lock-free atomic queues provide deterministic real-time performance without garbage collection pauses.

RustConcurrencyLock-Free

Mathisha

READ DEEP DIVE