AI-First Creative Production Workflow

We went from managing creative requests by email to running a system that other Actinver teams independently decided to copy.

Nobody asked for this. I mapped the full creative process, identified where everything was breaking down, designed the system to fix it, built it in Airtable, trained the team, and presented it to executives as a performance visibility tool. That last part changed how Marketing was perceived internally.

Company: Actinver — national investment bank with multiple business units generating continuous demand for audiovisual content: social media, CRM, DOOH, video, events, paid campaigns.

Team: 4 designers, 1 design lead, 2 content coordinators, 2 community managers, ~25 active requesters from Marketing and Growth. Eventually adopted across other teams.

When I mapped out the full creative process, the problem was clear: every single request ended up in one person's email.

The deeper issue was organizational. Marketing's output was invisible to executives because there was no system to make it visible. The team was doing a lot of work and nobody could see it.

I built a creative operations system with Airtable as the main platform, supported by Slack and Institutional Mailing. The entire process — from intake to delivery — is now structured, automated, and visible.

I built the entire Airtable architecture and automation logic myself. Then I ran two tracks simultaneously:

Internal — Onboarding sessions for Marketing and Growth teams, documentation shared across the whole company, feedback collection, and two major iterations based on how people were actually using it. v1 was intake and assignment. v2 added the AI validation layer and the executive dashboard after seeing that the biggest gap wasn't speed — it was visibility.

Organizational — I presented the system to executives as a performance visibility tool. That shifted how Marketing was seen — from a reactive service that was hard to measure, to a structured operation with clear metrics.

The system wasn't something we launched once and moved on from. We treated it like a product: versioned, iterated, and improved based on real feedback.

In the first two months, production capacity went up 60%. Delivery times dropped by 2–4 days. After-hours work caused by reactive overload basically disappeared.

The clearest signal that it worked: Operations and Digital Lab independently decided to replicate the system for their own teams. No training, no coordination from my side. They just saw what Marketing was doing and built their own version. That's the best validation a system can get.

Beyond capacity and speed, the system gave the team something it never had before: data. That visibility gave us the numbers to demonstrate Marketing's impact to leadership with hard evidence instead of perception.

The ops system solved the process problem. But there was still a bottleneck: the actual production of each asset still required a designer to start from scratch every time. So I built the next layer — a standalone AI-powered creative platform, migrated to Claude Code.

The first version was prototyped in Figma Make to validate the concept and interface quickly. Once the logic was proven, I migrated it to Claude Code — turning it from a prototype into a real production system.

Once Airtable registers a new approved request, the platform picks it up automatically. It reads the brief, pulls from NanoBanana Pro and free stock, generates the piece grounded in the Actinver brandbook, adapts it for every social media format, and delivers it back to Airtable — reviewed for visual quality and copy accuracy.

Since launching in early 2026, the platform has been continuously improved — a daily production system with measurable output that keeps getting better.