Pierre Chilès AI-Driven Creative Portfolio

AI-Driven Creative Portfolio

AI-Driven Creative Production

Music · Visuals · Narrative Systems

I design and produce original creative works using generative AI as a controlled production tool, not a gimmick.

My work sits at the intersection of art direction, prompt engineering, and narrative design.
Each project is conceived as a system: a clear intention, a defined aesthetic language, and a reproducible workflow.

I create:

AI-generated music tracks with precise stylistic constraints
Visual universes mixing illustration, photorealism, and surreal composition
Structured prompt frameworks enabling consistency, iteration, and scale

Every output is authored, curated, and directed.
AI executes — I decide.

This portfolio showcases selected experiments and productions where generative models are pushed beyond default outputs to deliver distinctive, intentional, and usable creative assets.

AFDIT

Promotional Visual & Video Creation for Professional Legal Events

I designed and produced a series of AI-generated visuals and promotional videos for major professional events organized by the Association Française du Droit de l’Informatique et de la Télécommunication (AFDIT).

The objective was to renew the visual language of legal and institutional communication without compromising credibility, clarity, or professional standards.

Using generative AI as a creative accelerator under strict art direction, I delivered:

Event teaser videos optimized for LinkedIn and professional networks
High-impact key visuals for conference promotion
Consistent visual identities aligned with each event’s thematic focus (cybersecurity, AI, digital law, data governance)

Each asset was conceived to balance:

Editorial seriousness expected by legal professionals
Visual disruption to capture attention in saturated communication channels
Narrative coherence between speakers, topics, and event positioning

All outputs were fully authored, curated, and validated, with particular attention to:

Readability and legibility in professional contexts
Stylistic restraint compatible with CSP+ audiences
Reusability across formats (social media, email campaigns, event pages)

This work demonstrates how generative AI can be integrated into institutional-grade communication, not as a novelty, but as a controlled production tool serving strategic messaging.

Breakfast Seminar on Software Intellectual Property

AI-generated promotional teaser for an AFDIT legal event on software IP protection.

Cyberpunk jazz jingle, controlled lipsync, and bold contemporary visual direction.

AI & Health Law Conference

AI-generated promotional teaser for an academic legal conference on AI and health law, organized at Université Paris-Panthéon-Assas.

Cyberpunk baroque jingle, stylized motion design, and bold contemporary visual direction.

Judge & Digital Law Conference

AI-generated short film created for an AFDIT legal conference on the role of judges in the digital era.

Contemporary music with Latin lyrics, neo-impressionist and surreal visual language, cinematic pacing.

A deliberate shift away from familiar AI aesthetics, using generative tools as a narrative medium rather than a visual signature.

Breakfast Seminar on Intangible Assets & M&A

AI-generated director’s cut for an AFDIT M&A audit breakfast conference.

AI-generated jazz song (Suno) with vocals and lipsync, shown here for its artistic and technical value; a simpler, more concise version was ultimately chosen for the official communication.

GDPR Breakfast Conference

AI-generated exploratory video concept for an AFDIT GDPR breakfast conference.

Rotoscoped, comic-inspired animation and AI-generated music, presented as an artistic and technical exploration; ultimately not selected for official communication due to positioning considerations.

Cybersecurity Law Conference

AI-generated promotional video for an AFDIT cybersecurity law conference.

Traditional visual language, classical voice-over, and AI-generated baroque music; a deliberately conservative approach that delivered strong audience engagement and high attendance.

Copyright & Generative AI Conference

AI-generated promotional video for a joint AFDIT × IRPI-N conference on copyright and generative AI.

Restrained yet bold visual direction using AI-transformed Paris and Brussels imagery, designed to align with the regulatory and legal sensitivity of the subject matter

Phenix League - Legal Data Space

AI-generated cinematic short film created for a private launch event introducing Phenix League,

a legal chatbot service built on a Legal Data Space architecture.

Conceived in an ultra-wide 21:9 cinematic format, the five-minute film was designed for immersive projection, reinforcing scale, continuity, and narrative depth during a closed, invitation-only screening.

The film adopts a mythological narrative framework - rebirth, alliance, and collective order - to visualize how fragmented legal data can reorganize into a coherent and governed system.

Through symbolic transformation, the Legal Data Space is portrayed as rising from fragmentation and re-emerging as a unified ecosystem.

Epic visual storytelling is paired with an AI-generated orchestral score, built as a reinterpretation of space-opera orchestral language. The music evokes familiar themes through AI-driven stylistic transformation, reimagining well-known cinematic motifs into an original score.

Designed as an interpretative and experiential layer, the film does not demonstrate features or interfaces.

Instead, it translates a complex legal and technical architecture into a shared narrative experience, aligning myth, governance, and collective intelligence.

Prompt Sonic

Prompt Sonic is a creative research project exploring new forms of storytelling made possible by AI-driven music and video.

On this channel, I publish a wide range of experimental formats to demonstrate intentional, authored uses of AI — deliberately standing apart from the growing wave of generic, low-effort “AI slop.”

2026

My AI-boosted music video greeting card for 2026 is a short-form audiovisual piece combining original music, generative visuals, and narrative intent.
Conceived as a modern greeting card, it blends personal footage - with cameo appearances by myself and family members - into a stylized visual language inspired by contemporary video art, echoing the playful, self-referential approach of Pierrick Sorin.

Monsieur Elephant

AI-generated music video telling the story of a young girl and her emotional bond with her elephant plush toy.

The narrative is carried by an original AI-composed song inspired by the musical language of children’s animated features, blending expressive melodies, theatrical dynamics, and narrative-driven vocals.

The musical direction deliberately echoes the codes of contemporary animated musicals — in the codes of contemporary animated musicals (original composition; no direct thematic reuse) — reinterpreted through a generative pipeline rather than imitation.

Produced using a hybrid workflow (Midjourney, Kling, Suno), the clip combines character-led storytelling, stylized visuals, and tightly synchronized music to create a cohesive, emotionally accessible universe for young audiences.

The video was selected as a finalist in the Music Video category of Project Odyssey Season 2 (February 2025), recognizing both its narrative clarity and the controlled integration of generative tools.

This project demonstrates how Prompt Sonic can translate child-oriented storytelling and musical emotion into a structured, reproducible AI-driven creative system.

Love Prompts

AI-generated music album exploring intimacy, vulnerability, and sensuality through generative composition.

Released in September 2025, Love Prompts investigates how emotional nuance, desire, and fragility can be expressed within AI-generated music without relying on parody or technical spectacle.

The album focuses on texture, pacing, and vocal presence, favoring subtle dynamics over maximalist production. Rather than showcasing the technology, the compositions aim to let emotion lead — using AI as a listening instrument, capable of hesitation, tension, and softness.

Each track is built from carefully structured prompts designed to shape: vocal intimacy and phrasing, harmonic warmth and restraint, and a sense of closeness more often associated with human-performed music.

Love Prompts functions both as a musical release and as a research object, questioning how generative systems can carry sensuality and emotional credibility when guided by precise artistic intent.

Muse Digitale

AI-generated music video created for “Muse Digitale”, track 9 of the album Love Prompts.

The piece is structured as a call-and-response dialogue between a male and a female voice, unfolding over an Afrobeats × Congolese rumba–inspired rhythm.

The song explores themes of love, betrayal, and infidelity, using vocal interplay rather than linear narration.

Each voice answers, challenges, or mirrors the other, turning the track into a musical confrontation where desire and doubt coexist.

Visually, the video blends multiple graphic inspirations into a stylized, non-realistic aesthetic.

Conceived as an art-first, suggestive yet non-explicit visual language, Muse Digitale relies on suggestion, rhythm, and emotional charge rather than explicit depiction, aligning sensuality with musical structure and narrative form.

Other Musical Projects

In addition to the works featured here, I maintain a broad catalog of AI-generated music projects released across multiple platforms.

Some tracks are published on my Suno profile as part of ongoing experimental workflows, while others are officially released on major streaming platforms under the artist name Prompt Sonic - in many cases, both.

Most releases are accompanied by AI-generated visuals, including cover artworks and looping video formats, conceived as integral extensions of the music rather than promotional afterthoughts.

Stylistically, this body of work spans a wide range of genres and moods.

Technically, it is supported by a highly specialized custom GPT developed to enable advanced, controlled audio outputs through structured prompting - prioritizing coherence, musical intent, and expressive precision over random generation.

Taken together, these projects reflect an approach to AI music creation based on method, iteration, and artistic authorship, where sound and image evolve together within a unified generative pipeline.

Page updated

Report abuse