Speech Recognition Engineer Resume Example

Lead Speech Scientist · Microsoft

Redmond, WA · 2021 - Present

• Led a team of 5 engineers in developing next-generation ASR models for Azure Cognitive Services, improving transcription accuracy by 15% across key languages.
• Designed and implemented a novel semi-supervised learning pipeline, reducing the need for human-annotated data by 40% and accelerating model iteration cycles.
• Optimized ASR inference latency by 20% on edge devices through model quantization and efficient beam search algorithms, enhancing real-time application performance.
• Collaborated with product management to define roadmap and delivered 3 major features impacting millions of enterprise users globally.
• Mentored junior engineers, fostering a collaborative environment that resulted in two patent filings related to robust speech feature extraction.

Senior Speech Recognition Engineer · Amazon (Alexa AI)

Seattle, WA · 2017 - 2021

• Developed and deployed deep learning-based acoustic and language models for Alexa, contributing to a 10% reduction in word error rate (WER) for targeted domains.
• Engineered scalable data processing pipelines for millions of hours of speech data, improving model training efficiency by 25% and reducing infrastructure costs by 15%.
• Collaborated with product teams to integrate new ASR features, resulting in a 5% increase in user engagement and satisfaction for voice assistant interactions.
• Conducted extensive A/B testing and experimentation to validate model improvements, leading to data-driven decisions that impacted millions of users globally.
• Published 2 papers at major ASR conferences (Interspeech, ICASSP) on novel methods for noise robust speech recognition.

Ph.D. Computer Science · Stanford University

Stanford, CA · 2012 - 2017

• Thesis: "End-to-End Deep Learning Models for Low-Resource Speech Recognition" with a focus on neural network architectures.
• Awarded University Fellowship for outstanding academic achievement and research potential.
• Maintained a GPA of 3.9/4.0 while conducting extensive research and teaching assistant duties.

M.S. Electrical Engineering · University of Washington

Seattle, WA · 2010 - 2012

• Specialized in Signal Processing and Machine Learning, with coursework in speech recognition fundamentals.
• Graduated Cum Laude for academic excellence in a rigorous engineering program.
• Achieved a GPA of 3.8/4.0, contributing to research on acoustic feature extraction.

Lead Embedded ASR Engineer · Apple Inc.

Cupertino, CA · 2020 - Present

• Led a team of 4 engineers in optimizing on-device ASR models for next-generation iPhone, reducing model footprint by 35% while maintaining 98.5% accuracy.
• Engineered and deployed low-latency inference pipelines for local voice commands on Apple Watch, improving response time by 20% and enhancing user experience.
• Developed novel neural network architectures specifically for low-power speech processing on custom silicon, resulting in a 15% reduction in power consumption for continuous listening.
• Implemented robust on-device wake word detection algorithms, achieving 99.8% accuracy with a false reject rate below 0.5% in diverse acoustic environments.

Senior Speech Recognition Engineer · Syntiant Corp.

Irvine, CA · 2017 - 2020

• Designed and implemented ultra-low-power ASR pipelines for edge devices using neural decision processors, cutting power consumption by 40% compared to traditional DSP solutions.
• Collaborated with hardware architects to optimize deep learning models for custom NXP chipsets, boosting inference speed by 2x for always-on voice applications.
• Contributed to the development of novel keyword spotting models, achieving 95% accuracy on limited-resource IoT devices with only 256KB of RAM.
• Enhanced model quantization and pruning techniques, reducing memory usage by 30% for embedded voice applications without significant accuracy degradation.

Speech R&D Engineer · Qualcomm

San Diego, CA · 2012 - 2017

• Optimized ASR algorithms for Qualcomm's Hexagon DSP, achieving a 25% improvement in processing speed and reducing latency by 15% for mobile devices.
• Developed and fine-tuned acoustic models for Snapdragon platforms, leading to a 10% Word Error Rate (WER) reduction in noisy environments for mobile ASR.
• Implemented real-time speech enhancement techniques using deep neural networks, improving voice clarity by 18% for in-car communication systems.
• Pioneered the integration of advanced Voice Activity Detection (VAD) with low-power modes, extending battery life by 5% in always-on voice assistant features.

M.S. in Electrical Engineering, specializing in Signal Processing · Carnegie Mellon University

Pittsburgh, PA · 2010 - 2012

• Thesis: "Deep Learning Architectures for Low-Resource Speech Recognition on Embedded Systems"
• Graduated with High Honors, GPA: 3.9/4.0

B.S. in Computer Engineering · University of Texas at Austin

Austin, TX · 2006 - 2010

• Concentration in Digital Signal Processing and Embedded Systems
• Magna Cum Laude, GPA: 3.8/4.0