Audio DSP Deployment Engineer
Hark
$120,000 - $300,000 / year
Need a reasonable accommodation to apply or interview? Contact us.
JobMinglr uses automated technology to recommend jobs based on profile information and job preferences. Match Score does not determine eligibility for a position, prevent a user from viewing or applying to a job, or make hiring decisions on behalf of an employer.
Description
About Hark
Hark is an artificial intelligence company building advanced, personalized intelligence. One that is proactive, multimodal, and capable of interacting with the world through speech, text, vision, and persistent memory.
We're pairing that intelligence with next-generation hardware to create a universal interface between humans and machines. While today's AI largely operates through chat boxes and decade-old devices, Hark is focused on what comes next: agentic systems that interact naturally with people and the real world.
To get there, we're developing multimodal models and next-generation AI hardware together - designed from the ground up as a single, unified interface for a new era of intelligent systems.
About the Role
We're looking for an experienced Audio DSP Deployment Engineer to own how our audio algorithms work on device. Our algorithm team prototypes and validates new audio processing on the desktop in MaxMSP, JUCE, Python, and MATLAB, but the on-device implementation is its own design problem: what runs where, at what precision, in what block size, and within what compute, memory, and latency budget. You will own that design, reworking and extending algorithms so they hold up in real time on the embedded DSP cores inside our products, tuning them against real hardware and real acoustics, and carrying them through to a shipping feature. This is hands-on work close to silicon with genuine ownership of the on-device audio chain, and direct influence on how the product sounds and listens.
Responsibilities
- Own the on-device design of the audio algorithm chain: what runs where, at what precision and block size, and within what compute, memory, and latency budget
- Take algorithms from desktop prototype in MaxMSP, JUCE, Python, and MATLAB to production C/C++ running in real time on the embedded DSP cores, re-architecting them where the target demands it rather than transcribing them
- Develop, tune, and extend algorithms directly on device, and propose new processing where measurement or listening says the current approach falls short
- Optimize for the target DSP architecture: vector and SIMD intrinsics, fixed-point versus floating-point trade-offs, memory placement and DMA, cache and tightly coupled memory behavior, block size and latency tuning
- Profile and budget performance, measuring cycle load, memory footprint, and end-to-end latency, and removing hot spots
- Validate against both the reference implementation and measured and perceived audio quality, and build the test fixtures and regression scripts that prove it
- Debug on-device audio issues using platform tuning, logging, and real-time trace tools
- Work with the algorithm, acoustics, and hardware teams to shape what the on-device chain should do, not only how it gets built
Requirements
- Strong modern C/C++ for real-time, resource-constrained embedded systems
- Solid DSP fundamentals: filtering, FFT and overlap-add processing, sample rate conversion, fixed-point arithmetic, and numerical precision
- Ownership of on-device algorithm design decisions: restructuring processing to fit real-time constraints, choosing precision and topology, and knowing when to redesign rather than port
- Comfort working from measurements and listening as well as from a reference implementation
- Demonstrated experience porting reference algorithms from MATLAB, Python, or desktop C++ to an embedded DSP or SIMD target while meeting cycle and memory budgets
- Fluency in an embedded toolchain: cross-compilation, on-target debugging, and profiling
- Experience with embedded audio frameworks and real-time processing pipelines on an SoC platform
- Ability to work directly with algorithm authors, acoustics, and hardware teams
Bonus Qualifications
- Experience developing audio processing modules within a vendor DSP audio framework on a mobile or consumer SoC platform
- Familiarity with MaxMSP and/or JUCE as prototyping environments
- Audio domain depth in echo cancellation, noise suppression, beamforming with MEMS microphone arrays, loudspeaker protection, or spatial audio
- SIMD optimization on ARM NEON or comparable vector units, RTOS concepts, and audio HAL or driver-level integration
- Acoustic measurement literacy (Klippel, SoundCheck, APx) and experience building automated audio regression tests
Compensation
The US base salary range for this full-time position is between $120,000 - $300,000 annually.
The pay offered for this position may vary based on several individual factors, including job-related knowledge, skills, and experience. The total compensation package may also include additional components/benefits depending on the specific role. This information will be shared if an employment offer is extended.
About this role
Hark is building multimodal AI systems paired with next-generation hardware, and this role owns how audio algorithms perform on the embedded DSP cores inside those devices. You'd take prototypes developed by the algorithm team in MaxMSP, JUCE, Python, and MATLAB—validated on desktop—and redesign them for real-time execution on constrained hardware. This means making hard trade-offs between precision, block size, compute budget, memory footprint, and latency; optimizing for the target DSP architecture using vector intrinsics and fixed-point arithmetic; and profiling cycle load and end-to-end latency to remove bottlenecks. You'd validate your work against both the reference implementation and actual audio quality, measured and perceived, and debug issues using platform tools and real-time tracing.
The role demands strong modern C/C++ for embedded systems, solid DSP fundamentals across filtering, FFT, sample rate conversion, and fixed-point math, and proven experience porting algorithms from high-level prototyping environments to resource-constrained targets while meeting hard performance budgets. You need to own the architectural decisions—knowing when to restructure rather than simply port—and work comfortably from measurements and listening tests alongside reference implementations. Experience with embedded audio frameworks, cross-compilation, on-target debugging, and profiling on SoC platforms is essential.
If you have depth in audio domains like echo cancellation, noise suppression, beamforming with MEMS arrays, or spatial audio, and hands-on experience with SIMD optimization on ARM NEON or acoustic measurement tools, you'd be well-positioned. This is hands-on work close to silicon with direct influence on how the product sounds and listens. The role is full-time, in-office in San Jose, with a salary range of $120,000–$300,000.
This role's local market on JobMinglr
Pay for this role
$120,000 to $300,000 per year
That is the employer's posted range, and where you land in it is usually decided in one conversation. Before that conversation, run the numbers: analyze an offer for this role.
Before you apply, worth reading
How JobMinglr reads this job
Every listing here is scored against your profile before you apply: skills overlap, experience level, location and work arrangement, each weighted and explained. You see the score and the reasons, not just a list. How the matching works.