Multilingual Video Source Normalization Pipeline
Design of a hybrid system (web tool + automations) to clean, match, and normalize multilingual video sources at scale, for industrial exploitation.
Intelligent automations, data pipelines, production workflows — I lay the technical foundations that let creative teams focus on what matters.





From creation to analysis, I work on the 4 pillars that drive a media company — and especially on their coherence.
Creation workflows, asset management, team coordination
Editing, validation, exports, AI automations
Multi-platform publishing, scheduling, syndication
Reporting, attribution, decision dashboards
Design of a hybrid system (web tool + automations) to clean, match, and normalize multilingual video sources at scale, for industrial exploitation.
Design of an internal system for calculating YouTube revenue shares at scale, intended to replace costly and heterogeneous external tools.
Design and launch of a SaaS platform for analyzing and managing social performance, dedicated to media companies and editorial teams.
Media has changed. One video becomes a short, a podcast, a Fast TV clip, 50 social posts... Without structure, this equation explodes.
1 production
Long-form video, documentary, show...
10 formats
Short, podcast, article, clip, story...
100+ publications
YouTube, TikTok, Fast TV, LinkedIn, Spotify...
The question is no longer "should we publish everywhere?" but "how to do it?"
Complete audit of your processes, tools and data flows.
Prioritization: what to automate, what to keep manual, what to stop.
Robust system design. Data pipelines, automations.
Long-term follow-up, training, skills transfer.
Building robust systems: agents, workflows, API integrations. From prototype to production.
Building agents to automate your recurring tasks
Automatic subtitling and translation of your content
Building business tools and process automation
Collecting, transforming and centralizing your data
Automated production of images, voices and assets
Connecting to your platforms (YouTube, Meta, TikTok...)
Python, n8n, Airtable, Make, dbt, BigQuery — the right tools for each problem.