Qualcomm
Website:
qualcomm.com
Job details:
About the Role
The role involves system-level validation and testing of Voice AI features, focusing on Always-On AI and Agentic AI use cases, while ensuring high performance and quality standards.
Responsibilities
- Voice AI Validation & System Testing
- Perform system-level validation and testing of Voice AI features, with emphasis on Always-On AI and Agentic AI use cases.
- Design, execute, and maintain comprehensive test plans for Voice AI pipelines including ASR, TTS, NLP, Translation, Summarization, and Language Models.
- Validate functional correctness, latency, power, memory, and long-run stability of Voice AI systems on embedded hardware (NPU, GPU, CPU).
- Maintain Performance dashboard
- Detect regressions
- Identify quality gaps
- Predict performance trends across releases
- Performance, Quality & Competitive Benchmarking
- Compare performance and perceptual quality of Qualcomm Voice AI solutions against:
- Internal reference implementations
- Customer baselines
- Competing open-source solutions in the market
- Conduct objective and subjective quality analysis for Voice AI features (accuracy, intelligibility, naturalness, latency perception, robustness).
- Drive validation with the goal to meet and beat market expectations, using measurable KPIs and user-experience metrics.
- Perform competitive benchmarking of open-source and proprietary models from a system-level quality and performance standpoint.
- Hardware Acceleration & System Integration Validation
- Validate HW AI acceleration paths, ensuring correct and efficient offload across NPU, GPU, CPU, and DSP.
- Evaluate end-to-end voice AI pipelines, which includes low power modes, and realtime inference behavior in batch mode and streaming mode.
- Verify correct interaction across application, framework, DSP, and hardware layers.
- Automation, Debug & Cross-Team Collaboration
- Develop and maintain automated test suites, regression frameworks, and evaluation pipelines for Voice AI and Agentic AI workflows.
- Analyze logs, traces, metrics, and dumps to perform root-cause analysis of functional, performance, or quality issues.
- Collaborate closely with R&D, Systems, Platform, Product, and Customer teams to drive fixes, improvements, and release readiness.
- Document test methodologies, KPIs, validation results, and competitive insights for internal and external stakeholders.
Qualifications
- Bachelor’s / Master’s / in Engineering, Electronics & Communication, Computer Science, or related fields.
- 2+ years of experience in Voice AI testing, Audio/AI system validation, ML inference evaluation, or related domains.
Required Skills
- Strong programming and scripting skills in C/C++ and Python, with focus on test automation, data analysis, and evaluation tooling.
- Experience in system-level testing and validation of ML or AI workloads on embedded platforms.
- Solid understanding of Voice AI domains: ASR, TTS, NLP, multilingual translation, and voice assistants.
- Hands-on experience validating ML inference performance and quality on NPU/GPU/CPU.
- Working knowledge of deep learning frameworks such as PyTorch, TensorFlow, ONNX from a testing and inference evaluation perspective.
- Understanding of ML architectures (Transformers, LSTM, GRU, diffusion models) for validation, benchmarking, and analysis.
- Experience validating model quantization, compression, and hardware acceleration techniques.
- Strong debugging skills for embedded systems, DSP pipelines, and AI accelerators.
- Experience with AI-assisted test data analysis, model-based evaluation, and KPI-driven benchmarking is a strong plus.
- Excellent communication and collaboration skills to work across global, cross-disciplinary teams.
Click on Apply to know more.