David MartinCreative Professional
01 · AI video

Building a promotional series with an AI-native production workflow.

A six-part promotional series for a technology solutions and managed IT services provider, developed, produced and edited with an AI-assisted workflow.

The brief

Create a polished promotional video series without relying on a conventional production model. Six videos, one series, one recognizable brand.

The creative idea

Humour up front, a serious point underneath. Each video takes one business risk and gives it an everyday visual metaphor: a goldfish and a baited hook for phishing (Don’t Take the Bait), an office move for a cloud migration (Moving Offices Without Boxes).

The visual direction

The same 3D animated look across all six, so each one is recognisable as part of the set. Series rules fixed before the first script: the look, the tone, the length, the ending.

The production challenge

Hold the look from clip to clip when every scene is generated. Keep six separately produced videos feeling like one series without repeating a single idea.

The technology

Clips generated in Google Flow from the same style and character references. Voice with ElevenLabs. Cut in CyberLink PowerDirector, so pacing and the closing call to action match across the series.

The workflow

Idea → Treatment → Visual references → Generation → Selection → Animation → Edit → Review → Final. The technology accelerated exploration. Creative direction determined what survived.

The result

Six finished promotional videos plus a two-part AI training series (The ELIZA Effect), each with its own metaphor, title and angle on the same promise: IT that is secure, reliable and easy to live with.

What I learned

AI doesn’t remove the creative director. It changes what the creative director can direct.

02 · Children’s media

Directing character-driven content across an 18+ year catalogue.

DJC Kids: an original children’s channel treated as a directed production for over a decade. Stories written, characters designed, animation directed, publishing run end to end.

The brief

Build an original kids’ property that can publish for years without losing its identity.

The creative idea

Character-driven stories with a consistent world: the same characters, the same voice, the same standard, upload after upload.

The visual direction

A locked series bible: model sheets for hand-drawn work, locked character profiles for AI work. One canon, every clip.

The production challenge

Publish at volume for a decade without drift. Move from hand-drawn 2D to AI-assisted 3D (Bob and Kale) without breaking character.

The technology

Hand-drawn 2D origins; AI-assisted 3D pipeline with locked references; YouTube as the distribution system.

The workflow

Story → character lock → production → review → publish. Every upload reviewed before release.

The result

261.5M+ lifetime views, 288K subscribers. Consistency beat virality, every year.

What I learned

Keeping a series on model is a publishing discipline first and a technology problem second.

03 · Brand

Building visual systems that keep a technology company recognizable.

Eleven years owning content, brand and proposals inside an established technology solutions and managed services provider.

The brief

Give a B2B technology company a recognizable visual identity across every touchpoint, and keep it coherent as the company grows.

The creative idea

One visual world: brand refresh, website rebuild, campaign systems, event presence, all speaking the same language.

The visual direction

A defined brand system applied everywhere: web, social, sales tools, events like MISA BC 2026, paid campaigns.

The production challenge

Keep dozens of assets per month on-brand with a small team. Make technical subjects visually engaging without dumbing them down.

The technology

Web platform rebuild, Google Ads, social publishing systems, AI-assisted production for volume work.

The workflow

Brand standards → templates → production → review. The system carries the consistency so the team can carry the ideas.

The result

A coherent brand across web, campaigns, events and sales materials, maintained for over a decade.

What I learned

Brand is a system, not a logo. The guidelines are only as good as the workflow that enforces them.

04 · Content

Turning complex technology into work people can actually understand.

Articles, sales tools and social content that make managed IT and cybersecurity clear without losing the substance.

The brief

Explain complex technology to non-technical buyers, consistently, at volume.

The creative idea

Everyday metaphors for technical problems. The same instinct as the video series: make the abstract visible.

The visual direction

Editorial clarity: clean layouts, consistent voice, visuals that explain rather than decorate.

The production challenge

Volume without dilution. Dozens of pieces a month, each one accurate, on-voice and genuinely useful.

The technology

Content systems with reusable structures; AI-assisted drafting with human review at every stage.

The workflow

Brief → draft → expert review → refine → publish. Accuracy is a creative constraint, not an afterthought.

The result

A content engine that feeds web, social, sales and campaigns from one consistent voice.

What I learned

Clarity is a creative decision. The best explanation is usually the simplest visual one.

05 · Proposals

Creative storytelling under real business constraints.

Proposal production treated as a directed creative system, not paperwork.

The brief

Win complex B2B work with proposals that are persuasive, accurate and on-brand, under real deadlines.

The creative idea

A proposal is an argument with a visual system. Structure it like a story: the client’s problem, the approach, the evidence, the ask.

The visual direction

Consistent proposal design language: the brand’s voice and visuals applied to high-stakes documents.

The production challenge

Complex RFPs demand substantial reading, searching, organizing and drafting, on a deadline, without errors.

The technology

An AI-assisted workflow: Extract → Understand → Map → Research → Draft → Review.

The workflow

Requirements → Knowledge → Strategy → Structure → Draft → Evidence → Review → Compliance → Approval. Each stage has a job, a source of truth and a quality standard.

The result

A complex RFP’s first draft moves from hours of manual work to roughly 30 minutes. Less time assembling information; more time strengthening the argument.

What I learned

Automation creates the capacity. Direction determines the value. The larger the production, the more important the system becomes.

06 · AI systems

Building the infrastructure behind scalable creative production.

The systems underneath the creative work: knowledge, workflow and approval infrastructure that lets production scale.

The brief

Make organizational knowledge usable for both people and AI, so production can scale without losing quality.

The creative idea

Build the system around the intelligence, not just the interface to it. A reliable reference layer everything else draws from.

The visual direction

Systems you can see: organized knowledge, clear workflows, visible approval gates.

The production challenge

Important information lives across documents, proposals, websites, presentations, emails and people’s heads. Generation is only useful if the underlying knowledge is trustworthy.

The technology

Interconnected company knowledge system feeding proposal, content and sales workflows.

The workflow

Source of truth → shared context → repeatable workflow → human approval layer.

The result

Knowledge that feeds production instead of sitting as a static archive. AI working from the organization’s actual language, services and history.

What I learned

The production system starts before the prompt.

The Lab

Experiments in creative technology.

Active exploration of where creative production is going. Each experiment asks one question.

Experiment 01

Character consistency

How do you maintain a recognizable character across multiple shots?

Experiment 02

Camera control

How do you separate camera movement from character generation?

Experiment 03

Multi-pass generation

What happens when you stop asking one model to do everything?

Experiment 04

AI cinematography

Can traditional cinematography concepts be translated into generative workflows?

Experiment 05

Synthetic production

What happens when locations, characters, lighting and props become computational?

Experiment 06

AI-directed editing

What parts of editorial decision-making can be systematized?

Experiment 07

Generative sound

How does AI change the relationship between image, music and sound design?

Experiment 08

Creative agents

Can an AI agent actually participate in a production without flattening the creative process?

More in the archive.

Treatments, storyboards, styleframes, and process work, available on request.

Get in touch →