Qualcomm is a world-leading provider of mobile, computing, XR, automotive, and IoT SoCs and software technologies. Our Snapdragon platforms deliver highly integrated CPU/DSP/NPU capabilities that enable premium on-device experiences with industry-leading power efficiency. 
 

 As a member of Qualcomm’s Windows Audio Software team, you will design and build next-generation on-device Voice AI capabilities on Windows-on-Snapdragon platforms—spanning low-power “always-on” sensing through interactive speech experiences. You will collaborate with cross-functional teams (hardware, DSP, systems, ML, and product) and work with global stakeholders across regions and sites to deliver robust, production-ready solutions.
 

Responsibilities

• Own feature development end-to-end for Voice AI capabilities on Windows, including wake-on-voice, on-device ASR, translation, transcription, and TTS. 

• Gather and refine technical requirements with internal and external partners; translate product goals into clear engineering requirements, architecture, and milestones. 

• Co-design and implement Windows audio + driver stack solutions that leverage Qualcomm SoC CPU/DSP/NPU capabilities while meeting strict latency and power targets for Voice AI experiences.

• Build or extend components across user-mode and kernel-mode boundaries (UMDF/KMDF), integrating with Windows audio frameworks and platform services as needed for Voice AI pipelines.

• Drive low-power audio architecture for “always-on” use cases, including concurrency and power-mode transitions across multiple audio features.

• Optimize and validate performance using Windows diagnostics and systematically debug complex multi-threaded and multi-processor issues across HW/SW boundaries.

• Communicate development status, risks, and mitigations to technical leads and stakeholders.

• Participate in roadmap planning and execution; contribute to process improvements that increase quality, scalability, and maintainability.

• Master's Degree in Engineering, Information Systems, Computer Science or related field. 

• 2+ years of industry experience in software engineering, with a focus in one or more of the following areas: Windows system software, device drivers, multimedia/audio, platform software, or performance/power-sensitive products.

• Proficiency in systems programming (C/C++ preferred) and ability to deliver production-quality code in complex codebases.

• Strong Windows development skills using Visual Studio, source control (Git-based workflows), and disciplined debugging practices.

• Solid understanding of Windows OS fundamentals relevant to system-level development: processes/threads, synchronization primitives, memory, I/O, and performance diagnostics.

• Experience developing or integrating Windows drivers using KMDF and/or UMDF, including build, deployment, and debugging fundamentals.

• Hands-on experience with multi-threaded programming, concurrency control, and performance/latency tuning.

• Proven ability to work cross-functionally and communicate effectively with distributed teams across regions/time zones.

• Windows audio ecosystem experience (any of the following): audio drivers, audio endpoints, WASAPI, audio effects/APOs, audio policy, power-aware audio scenarios, or end-to-end audio pipeline integration. 

• Demonstrated experience using AI-assisted development tools to improve productivity across the software lifecycle (requirements/architecture, implementation, debugging, testing, and documentation), with the ability to iterate on prompts and validate outputs against real code and requirements.

• Familiarity with mainstream AI coding assistants (e.g., GitHub Copilot / Microsoft Copilot) applied to practical engineering tasks such as code comprehension, change drafting, test scaffolding ideas, and documentation—while maintaining strong review and correctness discipline.

• Ability to create and share repeatable prompt patterns/templates for common engineering workflows and continuously refine prompts to increase specificity and accuracy.

• Experience with low-power / always-on audio and power-state transitions (e.g., Modern Standby constraints and wake/trigger flows).

• Familiarity with on-device voice activation building blocks and staged verification concepts (e.g., multi-stage keyword/wake approaches).

• Experience optimizing workloads across heterogeneous compute (CPU + DSP + NPU) and understanding tradeoffs across latency vs. power vs. quality.

• Exposure to Voice AI pipelines: ASR/TTS, speech translation, transcription, streaming/buffering strategies, and real-time constraints.