Automation · AI · Switzerland · Freelance

AI systems that save 5+ hours every week.

Senior Sound Technician specialised in automation and AI. I build email triage systems, lead automation, and audio pipelines for businesses that want to scale without hiring.

Let's Talk View Projects
About
Senior Sound Technician with experience in production, post-production and AI-powered audio workflow automation. I build systems that connect technical audio expertise with modern automation for studios, podcasters and agencies.
Alberto Gálvez
Results
45'
Saved every morning with the automated email triage system
−80%
Reduction in manual work for audio post-production
−5h
Per podcast episode compared to a manual production workflow
100%
Of leads and emails classified automatically, with zero human input
Services
01 —

AI Email Assistant

Automatic triage, priority classification and draft reply for every incoming email. Inbox under control with zero effort.

From CHF 350
02 —

AI Customer Support

AI assistant that answers FAQs, classifies requests and escalates complex cases to the human team. Available 24/7.

From CHF 500
03 —

Lead Systems

Automated client capture with forms, CRM, AI and intelligent real-time notifications.

From CHF 600
04 —

AI Podcast Production

Automated production pipeline: audio cleanup, transcription, show notes and publishing. For solo podcasters.

From CHF 490
05 —

Audio Workflow Automation

Post-production workflow automation for studios, agencies and production companies. Cleanup, EBU R128 normalisation and automatic QC.

From CHF 890
06 —

Landing Pages

Modern, minimal and conversion-focused websites. Premium digital presence for businesses that want to stand out.

From CHF 350
Case Study
01 —

AI Email Assistant

Intelligent system that monitors the inbox in real time, classifies each email by category and priority, generates an actionable summary, recommends the next step, and drafts a professional reply ready to send. The user reviews a structured dashboard every morning instead of reading raw emails.

Challenge
Manual email review consumed 45+ minutes every morning — no system to prioritize, triage, or draft responses automatically.
Solution
Automated classification pipeline using Make + Claude AI: every incoming email is categorized, summarized, and paired with a ready-to-send draft reply.
Result
45 minutes saved daily. 100% of emails classified without manual input. Zero high-priority leads missed since implementation.
MakeGmail APIClaude AI Google SheetsWebhook
ENTRADA TRIGGER AI PROCESS ALMACÉN SALIDA Gmail Incoming email Make Webhook trigger Claude AI Clasifica + redacta Google Sheets Log + prioridad Dashboard Resumen diario
View full case study — technical flow, tiers and metrics
System Flow
01
Input Trigger
Make detects every new email in Gmail in real time. Extracts sender, subject and message body para su análisis.
MakeGmail API
02
AI Classification & Analysis
Claude AI processes the email and returns a structured JSON: category (Sales, Support, Admin, Spam), prioridad (Alta / Media / Baja), resumen ejecutivo en 2 líneas y acción recomendada.
Claude AIJSON output
03
Suggested Reply Drafting
In the same step, the AI generates a professional reply ready to send, adapted to the sender's tone and email type.
Claude AIPrompt engineering
04
Dashboard Logging
Google Sheets logs every email with all fields: date, sender, subject, category, priority, summary and draft. The user reviews the dashboard every morning.
Google Sheets
Live Dashboard — System Data
Remitente
Asunto
Category
Prioridad
Recommended Action
Carlos Mendoza · LogiTrans S.A.
Urgent: Question about web development budget
Ventas
Alta
Book a 10-min call tomorrow to discuss requirements and pricing
Fwd: Carlos Mendoza · LogiTrans S.A.
Re: Urgent — delivery deadlines and availability
Ventas
Alta
Contact immediately to schedule the call and confirm availability
Service Tiers
Tier 1 — Basic Triage
CHF 350–500
One-time setup
  • Gmail + AI classification
  • Dashboard Google Sheets
  • 4 configured categories
  • 2 semanas soporte
Tier 3 — CRM Integration
from CHF 1,200
custom project
  • Multi-cuenta
  • Existing CRM integration
  • Categorys y reglas custom
  • Onboarding video call
System Results
45'
Saved every morning on inbox review
100%
Of emails classified without human input
0
High-priority leads missed since implementation
Featured Project
02 —

AI Lead Automation System — Aeschi Studio

Intelligent lead capture system connected to AI, CRM, Google Sheets, Telegram and advanced automation via Make.

Challenge
Leads arriving via web form were handled manually — no scoring, no instant follow-up, no CRM record. High-value contacts slipped through the gaps.
Solution
End-to-end lead automation: form submission triggers Gemini AI scoring, instant Telegram alert with lead summary, and automatic CRM entry in Google Sheets.
Result
Every lead scored, logged, and notified in under 60 seconds. Response time reduced from hours to minutes. Full contact history available from day one.
MakeGemini AITelegram Google SheetsWebhookGmailGitHub Pages
ENTRADA TRIGGER SCORING IA ALERTA CRM Formulario Incoming lead Make Webhook trigger Gemini AI Califica + resume Telegram Notification <60s Google Sheets Registro auto
Completed Project
03 —

AI Booking System — JotaGrind Studio

Premium booking landing page for a high-end tattoo studio. End-to-end system: intelligent intake form, lead analysis with Gemini AI, instant Telegram alerts and automatic CRM in Google Sheets. Deployed in production and delivered to a real client.

Challenge
A high-end tattoo studio was managing bookings via DMs and manual emails — no way to filter serious clients from casual inquiries or track pipeline status.
Solution
Premium booking landing page with intelligent intake form. Each submission analyzed by Gemini AI to score lead quality, trigger Telegram alerts, and build a live CRM in Sheets.
Result
Full booking system deployed and delivered to a real client. Studio owner receives instant AI-scored lead summaries with zero manual data entry.
MakeGemini AITelegram Google SheetsWebhookVercelHTML / CSS / JS
CLIENTE TRIGGER SCORING IA ALERTA CRM ENTREGA Landing Reserva Make Webhook Gemini AI Scores lead Telegram Notification Google Sheets Registro auto Cliente Live
Active Service
04 —

AI Podcast Production System

End-to-end automated system that transforms raw audio into ready-to-publish editorial content. The podcaster records, uploads a file, and within 15 minutes receives processed audio, transcript, show notes, alternative titles, social quotes and timestamped chapters. Built for solo podcasters without a production team.

Challenge
Solo podcasters were spending 5+ hours per episode on audio cleanup, transcription, show notes, and social content — a production bottleneck that limited output frequency.
Solution
Fully automated pipeline: upload raw audio → Auphonic cleans and normalizes → Whisper transcribes → Claude generates all editorial content → Telegram delivers the complete package.
Result
5 hours per episode reduced to under 15 minutes. 6+ deliverables generated automatically: processed audio, transcript, show notes, titles, quotes, and chapter markers.
MakeAuphonic APIWhisper API Claude APIAbleton LiveTelegramGoogle Drive
ENTRADA AUDIO TRANSCRIPCIÓN CONTENIDO IA ORQUESTA ENTREGA Audio bruto Subida archivo Auphonic Limpia + normaliza Whisper Transcription Claude AI Notes + titles Make Orchestrates flow Telegram Paquete listo
View full case study — technical flow, tiers and metrics
System Flow
01
Input Trigger
Make detects the new file in /ingest/ on Google Drive. Verifies naming convention and logs the entry in the production log.
MakeGoogle Drive
02
Audio Cleanup & Normalisation
Auphonic API applies noise reduction, hum elimination and loudness normalisation to -16 LUFS (streaming standard). The clean file is saved to /processed/.
Auphonic APIEBU R128 · -16 LUFS
03
Transcription with Timestamps
Whisper API transcribes the episode with timestamps per paragraph. The .txt is saved to /transcripts/ and passed to the next module.
Whisper APINeutral Spanish
04
AI Content Generation
The transcript is processed with the Claude API. Returns JSON with 3 titles, platform description, show notes, 5 social quotes and a chapter list with timestamps.
Claude APIJSON output
05
Logging & Notification
Google Sheets logs the episode with all metadata. Telegram Bot sends the client a preview of the generated content for review before publishing.
Google SheetsTelegram Bot
06
Automatic Publishing (Tier 2/3)
After client approval via Telegram, Make uploads the episode to RSS.com or Buzzsprout with title, description and chapters. No manual intervention.
RSS.com APIBuzzsprout
Ableton Live — Recording Template included in onboarding
Pre-configured Live Set for the client

The client receives a ready-to-open Live Set for recording. Ensures audio enters the pipeline in condiciones óptimas sin necesidad de conocimientos técnicos de audio.

Pistas
Host + Guest · Mono
Cadena
EQ · Gate · Comp · EQ
Export
WAV 24bit · 44.1kHz
Naming
EP001_HOST.wav
Service Tiers
Tier 1 — Ready to Record
CHF 490–690
One-time setup
  • Full Make workflow
  • Auphonic configurado
  • Ableton Live Set
  • Telegram Bot activo
  • 2 semanas soporte
Tier 3 — Full Studio
from CHF 1,200
custom project
  • Email + social integration
  • Approval workflow
  • Onboarding video call
System Results
~15'
From raw audio to review-ready content
−5h
Per episode vs. manual workflow
6+
Deliverables generated automatically
Active Service
05 —

Audio Post-Production Automation Consultant

Pipeline automatizado de limpieza, procesado y entrega para estudios de doblaje, agencias audiovisuales y estudios de grabación. El sistema clasifica los archivos entrantes, separa fuentes, normaliza según el estándar de cada destino y ejecuta control de calidad automático. Criterio de Técnico Superior de Sonido integrado en cada parámetro del sistema.

Challenge
Post-production studios were processing audio files manually — no consistent loudness standards across deliverables, no automated QC, and 80% of time spent on repetitive technical tasks.
Solution
Automated processing pipeline: ingest → source separation via Lalal.ai → Auphonic normalisation per destination standard (Broadcast / Streaming / Dubbing) → FFmpeg QC → Telegram delivery report.
Result
80% reduction in manual processing time. 100% consistent technical output per file. Zero loudness or True Peak errors on delivery since implementation.
MakeAuphonic APILalal.ai API FFmpegWhisper APIEBU R128Telegram
ENTRADA SEPARACIÓN NORMALIZACIÓN QC AUTO ORQUESTA INFORME Audio bruto Clasificación Lalal.ai Separa fuentes Auphonic EBU R128 FFmpeg True Peak + QC Make Orquesta flujo Telegram Entrega OK
View full case study — technical flow, tiers and metrics
Pipeline de procesado
01
Ingest & Classification
Make detecta archivos nuevos, analiza nombre y metadata para clasificar el tipo de contenido y enrutar al pipeline correcto. Registra entrada en Sheets.
MakeGoogle DriveGoogle Sheets
02
Source Separation
Lalal.ai API separa voz, ruido de fondo y música en stems independientes. Limpiar cada elemento por separado da resultados significativamente mejores que limpiar el mix completo.
Lalal.ai APIStems separation
03
Cleanup & Normalisation by Destination
Auphonic aplica noise reduction, hum removal y de-essing. Loudness ajustado según destino: Broadcast −23 LUFS · Streaming −16 LUFS · YouTube −14 LUFS · Doblaje −18 LUFS + True Peak −1 dBTP.
Auphonic APIEBU R128True Peak
04
Automated Quality Control
FFmpeg webhook verifica LUFS final, True Peak y duración vs original. Si el archivo no pasa el QC, alerta inmediata vía Telegram con el motivo específico.
FFmpegTelegram Alert
05
Automatic Subtitles
Whisper API genera .srt y .vtt para proyectos de doblaje y localización, y .txt con timestamps para edición.
Whisper API.srt · .vtt · .txt
06
Packaging, report and delivery
Make organiza la carpeta de entrega. Genera informe técnico con LUFS, True Peak, herramientas y tiempo de procesado. Telegram notifica al técnico con el resumen completo.
Google DriveInforme QCTelegram Bot
Why a Senior Sound Technician
🎚️
Loudness parameters with professional judgement
EBU R128, −16 LUFS y −14 LUFS no son lo mismo. Cada destino tiene su estándar. El sistema aplica el correcto automáticamente para broadcast, streaming, doblaje o YouTube.
🔬
Source Separation antes de limpiar
Limpiar el mix completo da peor resultado que limpiar la voz sola y recombinar. Este orden de operaciones solo lo define alguien que entiende el audio a nivel técnico.
QC with real metrics
El sistema verifica True Peak, LUFS integrado y duración. Son las mismas métricas que piden broadcasters y distribuidoras en sus especificaciones técnicas de entrega.
🤝
Technical conversation as equals
Phase coherence, artifacts del de-noiser, limitaciones del compresor en ciertos materiales. Con el técnico del estudio se habla de tú a tú. Eso genera confianza y reduce fricción.
Service Tiers
Tier 1 — Basic Pipeline
CHF 890–1.200
One-time setup
  • Workflow for 1 content type
  • Auphonic configurado
  • Automatic QC + alerts
  • 3 semanas soporte
Tier 3 — Mantenimiento
CHF 290–490
por mes
  • System monitoring
  • Parameter adjustment
  • Nuevos pipelines
  • Soporte Telegram
System Results
−80%
Manual processing time per project
100%
Technical consistency across project files
0
Deliveries with loudness or True Peak errors
How I Work
01

Initial Consultation

We talk for 30 minutes to understand your business, your current processes and where the biggest bottleneck is. No commitment required.

02

Design & Build

I build the system from scratch: automations, integrations and workflows tailored exactly to your case. I keep you informed at every step.

03

Delivery & Support

I deliver the working system, walk you through how to use it, and remain available for post-delivery adjustments. Your investment does not disappear on day one.

Contact
Got a project in mind?
Let's talk.

Available for automation, AI and digital systems projects. I reply within 24 hours.