SIGNAL_ARRAY // qtvo.dev LISTENING

ARRAY:19CH  ·  DEPTH:1999m  ·  RANGE:5000m  ·  SR:48kHz  ·  SRC:SIM

Quoc Thinh
Vo

>

full-stack engineer  ·  ci/cd & cloud platform  ·  ai/ml researcher  ·  array signal processing

uses your mic · audio never leaves this page
▼

> whoami

Full-stack software engineer building AI tooling, cloud platform, and the CI/CD that ships it. By night, an AI/ML researcher teaching models to hear the ocean.

I've shipped cloud-native software end to end across the tech, medical-device, and pharmaceutical industries: front-end, to APIs, to the infrastructure underneath. I care about the whole life cycle, developer productivity, platform and infra reliability, and CI/CD pipelines done right. On the research side, I publish on machine learning for audio, acoustics, and signal processing.

Off the clock: mathematics rabbit holes, building my own desktop, assembling mechanical keyboards, and soccer with friends. Powered by coffee. Happy clacking. ☕

  • Python
  • JavaScript
  • React
  • Node.js
  • PostgreSQL
  • AWS
  • GCP
  • Docker
  • Kubernetes
  • Terraform
  • CI/CD
  • PyTorch
  • C/C++

> systemctl status engineering

  • swe-roku.service active (running)

    Roku Inc. · Jun 2023 → present

    Building AI tooling, platform, and infrastructure: cloud-native services, developer productivity, and CI/CD you can trust.

  • research-imaple.service active (running)

    iMaPLe Lab, Drexel University · Sep 2022 → present

    Machine learning for audio, acoustics, speech, and signal processing.

  • swe-moberg.service completed

    Moberg Analytics · Sep 2022 → Jun 2023

    Worked on the CONNECT App, a cloud platform for neurocritical care: built core paginated back-end APIs and reusable, dynamic React front-end components.

  • swe-intern-roku.service completed

    Roku Inc. · Jun 2022 → Sep 2022

    Grafana + PromQL dashboards for Jenkins monitoring with Slack alerting; provisioned dashboard management and data scrapers on Kubernetes, GitLab CI, AWS, Terraform, Bash, and Python.

  • swe-coop-moberg.service completed

    Moberg Analytics · Mar 2022 → Jun 2022

    Patient de-identification tooling (Python, Bash, IBM Cloud) and an HDF5 converter that grew supported data types from 1 to 4+; co-authored a poster used in UCSF clinical trials.

  • cloud-coop-bms.service completed

    Bristol Myers Squibb · Mar 2021 → Sep 2021

    Automated licensing management with React, Node.js, SCIM, LDAP, and AWS; shipped and documented full-cycle, scalable software via Infrastructure as Code; secured REST APIs with AWS API Gateway and SSO.

> analyze --deep featured transmission

Teaching machines to localize sound in the deep ocean, and chasing models that generalize to seas they've never heard.

// the problem

Pinpointing where a sound came from underwater is brutally hard. The ocean is a warped, noisy channel, and a model tuned to one environment falls apart in another. Range and depth estimates that hold in one sea simply don't transfer to the next.

// the approach

Two architectures, two philosophies. ACA-Net is a multi-branch network that fuses GCC-PHAT and mel-spectrogram features, extracting several views of the signal at once. SA-Net works straight from the raw waveform, pairing CNN-based spiking (neuromorphic) layers with Conformer-style attention. The shared goal is domain adaptation: getting a model to hold up in oceans it wasn't trained on. Early zero-shot tests are encouraging, with evaluation ongoing across more datasets and domains.

// the result

Two peer-reviewed IEEE publications, ACA-Net (EUSIPCO 2025) and SA-Net (ICASSP 2026), with the full code open-sourced: training, feature extraction, and everything in between. And if you just want the SA-Net model without the codebase, it installs straight from PyPI as the sanet package.

SA-Net · ICASSP 2026 → SA-Net code → ACA-Net · EUSIPCO 2025 → sanet on PyPI →

> ls projects/ 5 contacts detected

  • CONTACT 001 · ICASSP 2026 · FEATURED

    Spiking Attention Network (SA-Net)

    A hybrid neuromorphic model for underwater acoustic localization that explores zero-shot generalization to unseen environments. Runs straight from the raw waveform, pairing CNN-based spiking layers with Conformer-style attention. Shipped as the sanet package on PyPI.

    PyTorch · Spiking Neural Nets · Conformer Attention · Raw Waveform

    code → paper → PyPI →

  • CONTACT 002 · AVSS 2026

    Underwater Source Detection & Classification

    Signal-based surveillance for underwater sources: audio dataset curation and cross-domain evaluation for detecting and classifying sounds across different datasets and ocean environments.

    PyTorch · Source Detection · Classification · Cross-Domain

    code → paper →

  • CONTACT 003 · EUSIPCO 2025

    Adaptive Control Attention Network (ACA-Net)

    A multi-branch network for underwater acoustic localization and domain adaptation, fusing GCC-PHAT and mel-spectrogram features for richer multi-feature extraction. Evaluated on the SwellEx-96 experiment.

    PyTorch · Multi-Branch · Generalized Cross-Correlation with Phase Transform and Mel-Spectrogram · Domain Adaptation

    code → paper →

  • CONTACT 004 · MLSP 2023

    Hierarchical Audio Transformer

    Underwater acoustic signal classification with a hierarchical audio transformer that stays robust to noisy input. Published at IEEE MLSP 2023 (Rome).

    PyTorch · Audio Transformer · Underwater Acoustics · Classification

    code → paper →

  • CONTACT 005 · DCASE2024 · TASK 3

    Sound Event Localization & Detection

    Submission to the 2024 IEEE AASP Challenge (DCASE2024), Task 3: sound event localization and detection evaluated in real spatial sound scenes, with source distance estimation.

    SELD · Spatial Audio · DCASE2024 · Deep Learning

    paper → results →

> cat cv.log

publications

more publications on Google Scholar →

talks & conferences

  • 2026 22nd IEEE International Conference on Advanced Visual and Signal-Based Systems (AVSS) Lecce, Italy
  • 2026 51st IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) Barcelona, Spain
  • 2025 33rd European Signal Processing Conference (EUSIPCO) Palermo, Italy
  • 2024 Engineering Tooling Conference San Jose, CA
  • 2024 Developer Productivity Engineering Summit San Francisco, CA
  • 2023 IEEE 33rd International Workshop on Machine Learning for Signal Processing (MLSP Conference) Rome, Italy

> ls education/

degrees

  • M.S. Computer EngineeringDrexel University · Philadelphia, PA
  • B.S. Computer EngineeringDrexel University · Philadelphia, PA
  • High School DiplomaLe Hong Phong High School for the Gifted · Ho Chi Minh City, Vietnam

certifications

  • Terraform AssociateHashiCorp · Sep 2023
  • AWS Solutions Architect AssociateAWS · Oct 2021
  • AWS Cloud PractitionerAWS · Jul 2021
  • AWS Fundamentals: Building Serverless ApplicationsCoursera · Sep 2020
  • Jira FundamentalsAtlassian University · Mar 2022

> establish-comms