Generative AI Designer Job Description

Use this as a good starting place for your next hire. Make sure to tailor it to your stack, content surface area, and AI maturity.

A starting template. Tailor to your team's craft standards, tooling, and brand requirements.

Why this role exists

The work has to hold up. That's the bar.

Generative tools have changed how much of our visual output can be produced and at what speed. They have not changed our standards for what gets shipped. The brand still needs to feel like the brand. The work still needs to look like it came from us. The role exists to make sure scale doesn't come at the cost of craft.

You'll work across image, video, and motion using tools like Midjourney, Runway, and Adobe Firefly, and you'll integrate generative work into existing pipelines built around platforms like Figma. You'll partner with brand, marketing, and product to define when generative tools belong in the workflow, build prompt and asset systems that keep output on-brand, and ship work that holds up against the rest of the design team's craft.

Success is measured on the volume, quality, and consistency of work that ships, and on how reliably others can build on the systems you create.

What you'll be doing

• Designing and producing generative imagery, video, and motion across marketing surfaces: campaigns, web, social, lifecycle, brand, product.
• Building and maintaining the visual prompt library, asset system, and brand-conditioning workflows. Versioned and discoverable.
• Defining when generative tools should and shouldn't be used. Partnering with brand on visual identity guardrails.
• Evaluating new models and tools, running controlled tests, deciding what enters the team's stack.
• Working cross-functionally with brand, marketing, content, motion, and product designers. Pairing generative output with human-led design where appropriate.
• Building small repeatable workflows (asset variation, image-to-image upscaling, controlled video generation) that other designers and marketers can use safely.
• Staying close to the model and tooling landscape and bringing options to the team with clear tradeoffs.
• Leading the response to risks: IP and likeness exposure, accidental brand drift, accessibility, disclosure.
• Mentoring designers and marketers on generative craft, prompting for visual systems, and editorial review.

What you'll bring

The non-negotiables:

• 5+ years of design experience across digital, brand, motion, or production design, with at least one year shipping production work using generative tools.
• A portfolio of generative work that holds up next to the rest of your design portfolio. Specific tools, specific decisions, specific outcomes.
• Strong visual craft fundamentals: composition, color, type, motion, brand systems.
• Fluency with leading generative platforms across image, video, and motion. Comfortable evaluating new tools quickly.
• Working knowledge of design and production pipelines (Figma, Adobe Creative Cloud, motion tools, asset management).
• Comfort with structured prompt craft for visual work and the ability to teach it.
• Cross-functional collaboration. You can work with brand, marketing, content, motion, and product designers without friction.

What helps but isn't required:

• Hands-on experience with image-to-image, ControlNet-style techniques, or model fine-tuning on brand assets.
• Background in motion, 3D, or VFX.
• Experience integrating generative tools into ad production pipelines at scale.
• Familiarity with model APIs and the Hugging Face model ecosystem.
• Background in regulated or risk-sensitive industries.

What success looks like

By day 30. You've audited the current visual production workflow, generative tooling, and brand assets. You've talked to brand, marketing, content, and design partners. You've identified the highest-leverage gaps.

By day 60. Visual prompt library v1 is shipped, along with the brand-conditioning workflow and one repeatable production workflow (campaign asset variation is a common starting point). A quality rubric and baseline review are in place.

By day 90. Tooling and workflows are rolling out to the broader team. The first quarterly report on generative output — volume, quality, consistency, impact — is published.