Patrick Marmaroli — Audio Signal Processing, Web Development & ML Specialist

About

Signal processing engineer with a PhD from EPFL and 15+ years of experience building audio ML systems — from acoustic sensing prototypes to production-grade pipelines at Microsoft. I combine deep expertise in DSP, machine learning, and full-stack web development to turn complex problems into working software. Beyond audio, I design and build production web platforms for enterprise and multi-stakeholder clients. Currently a Senior Audio Engineer at Microsoft (Teams audio quality), Co-Founder of Vocametrix, a speech analysis API and interactive speech therapy games platform, and Co-Founder of Hidacs Sàrl, a custom software and R&D consultancy.

Looking for custom software development?

Through Hidacs Sàrl, I offer custom software engineering ("ingénierie logicielle sur mesure") from prototype to deployment — SaaS products, web applications, and complex business systems (ERP, approval workflows, document logistics) — alongside R&D consulting in audio signal processing and ML.

Visit hidacs.ch

Looking for speech processing APIs or voice-based medical applications?

Through Vocametrix, I offer a professional speech analysis API for developers — real-time speech recognition, pronunciation assessment, and voice quality metrics across 30+ languages, with clinical-grade accuracy for speech-language pathology (orthophonie) and other voice-related medical applications.

Visit vocametrix.com

Experience

Jan 2022 — Present

Senior Audio Engineer

Microsoft · Tallinn, Estonia

Sound quality engineering for Microsoft Teams — audio processing pipelines, quality metrics, and signal analysis at scale.

Oct 2023 — Present

CEO & Co-Founder

Vocametrix OÜ · Tallinn, Estonia

Voice-controlled games and audio analysis tools for speech therapy. Web platform combining pronunciation assessment, pitch tracking, and gamified exercises, backed by a speech processing API suite serving both the platform and external developers.

Feb 2016 — Present

Founder

Hidacs Sàrl · Lausanne, Switzerland

R&D consulting in signal processing, machine learning, and web development — feasibility studies, proof-of-concepts, and prototyping for speech, industrial, environmental, and health applications, alongside full-stack web platform development for enterprise and multi-stakeholder clients. Delivered ERP, e-signature, and timesheet portal systems for multi-partner organizations.

Oct — Dec 2021

Audio Signal Processing Consultant

Onfido

Development of an anti-spoofing technology using ultrasound for identity verification liveness detection.

Jul 2020 — Sep 2021

Research Engineer (part-time)

HEPIA (HES-SO Geneva)

EARSENS project — privacy-preserving indoor activity detection using electrodynamic loudspeakers and AI, without microphones. Co-authored publication in Applied Acoustics (January 2023).

May 2016 — Aug 2021

Audio Department Manager

HED Technologies · Geneva, Switzerland

Designed and implemented signal processing algorithms (echo cancellation, noise reduction, voice command recognition) and contributed to headphone acoustic design. Led R&D efforts growing the Audio Department from scratch to a 4-engineer team, steering algorithm development from prototype to validated product.

Sep 2014 — Jun 2016

Project Manager, R&D Acoustics Department

PRONA SA · Yverdon, Switzerland

Developed custom Matlab tools and data processing pipelines for building and environmental acoustic analysis. Conducted normative acoustic measurements and managed multidisciplinary acoustic projects for industrial, institutional, and private clients.

Mar 2013 — Oct 2014

Post-Doctoral Fellow

EPFL · Lausanne, Switzerland

Designed audio signal processing algorithms and conducted laboratory and field experiments within European research projects. Collaborated with industrial partners including Phonak (hearing aids), FIAT, and ARMASUISSE. Supervised interns and PhD students.

Education

2008 — 2013

PhD in Acoustics & Audio Signal Processing

Swiss Federal Institute of Technology (EPFL) · Switzerland

Bimodal sound source tracking applied to road traffic monitoring.

2007 — 2008

Master Thesis in Signal Processing

Université du Sud Toulon-Var · France

Signal processing, psychoacoustics, and trajectography.

2005 — 2007

Software Engineering Degree

ISEN Toulon · France

2003 — 2005

Classes Préparatoires (Physics & Engineering)

Lycée Lachenal · Argonay, France

Technical Skills

Audio & DSP

MATLABWeb Audio APIFFT AnalysisGCC-PHATHRTFWaveSurfer.js

AI & ML

PyTorchScikit-learnOpenAI GPT-4YAMNetAzure Cognitive ServicesCustom CNNsDeep Learning Toolbox

Languages

PythonJavaScript / TypeScriptC#MATLABSQLPowerBI

Web & Cloud

ReactNext.jsNode.js / ExpressC# / .NETAzure App ServicePostgreSQLpnpm MonoreposDockerGitHub ActionsCI/CD

Security

AES EncryptionPrivacy-Preserving AIGDPR Compliance

Languages

French

Native

English

C1 Advanced

Selected Projects

Select a category above to explore projects

Enable JavaScript for the interactive view. Below is the full project list.

Audio Technology

Content Creation & Media

Audio Comparison Tool: A web application for scenarios where comparative listening is required both for individual and collaborative use. — no public repo · read technical details
AI-Powered YouTube Thumbnail Generator: AI-powered thumbnail generation platform using DALL-E 3 with multi-step wizard comprehensive emotional tone analysis and target audience optimization. — no public repo · see the technical implementation details
ImmoGen – AI Real Estate Market Reports: An AI-powered real estate market analysis report generator using LLM agents with section-by-section iterative refinement professional PDF output and Stripe payment integration., — visit the web app

Developer Tools

GCC-PHAT Time Delay Estimator for Stereo Channels – Executable & C# Source Code: A command-line tool for estimating the time delay between the two channels of a stereo signal using the Generalized Cross-Correlation Phase Transform (GCC-PHAT) technique — access the repository · read the Linkedin article
AES file encryption and decryption: A versatile command-line tool for AES file encryption decryption and key generation., — access the repository
Consulting Finder: A B2B prospecting tool that scrapes web sources for consulting leads scores them and generates outreach messages via an LLM API with a CRM pipeline to track status., — no public repo
Vibe Language (VL): A token-efficient programming language that compiles to Python JavaScript TypeScript C and Rust — designed to reduce AI coding costs by automatically converting code into a more compact syntax before sending it to an AI model., — access the repository

Education & Learning

Fitness & Wellness

Food & Lifestyle

ChefFantastique: ,AI-assistant that generates custom recipes. — read technical details

Healthcare & Accessibility

Other

Security & Privacy

EARSENS - Privacy-Preserving Indoor Activity Detection: An innovative acoustic sensing system using electrodynamic loudspeakers and AI to detect indoor events without microphones ensuring privacy while monitoring occupancy room changes and security events., — see project details
Explosive Event Acoustic Detection and Classification: Advanced real-time audio analysis system for detecting localizing and classifying acoustic events using machine learning and signal processing with multi-channel directional analysis sound level monitoring and YAMNet-based event classification. — no public repo · read technical details