Divyesh

Embedded Software Engineer

Qualcomm • Full-time

Mar 2025 - Present • 1 yr 3 mos

• Designed and implemented ML-based application traffic classification feature in WLAN to enable intelligent Quality of Service (QoS), improving real-time traffic prioritization and power efficiency. • Worked on WLAN (Wi-Fi) driver development focusing on performance optimization, stability, and low-level driver enhancements using C across embedded Linux systems. • Debugged and resolved complex kernel and driver-level issues including packet drops, DMA unmap mismatches, memory leaks, and regression failures impacting system stability. • Performed crash and performance debugging using T32, identifying root causes such as buffer overflows, synchronization issues, and packet delay scenarios leading to device instability. • Analyzed and resolved customer and validation-reported issues through deep log analysis, ensuring production-ready fixes and improved WLAN reliability. Generative AI & Automation: • Built internal automation tooling for WLAN log analysis and issue triaging, reducing debugging turnaround time from 3–4 hours to 15–20 minutes. • Developed an MCP server with code indexing and tagging support for efficient symbol/API retrieval, minimizing full-file parsing and reducing LLM inference cost. • Built tooling for large-scale codebase analysis and automated feature documentation generation including flow diagrams, sequence diagrams, DSA relationships, dependency mapping, and feature-level analysis. • Worked extensively with GenAI models including Claude, Gemini, and GPT for code analysis, design optimization, code improvement, and engineering productivity automation.