Dionisis G.
Available · --:-- EET

(AI Designer — Greece, open to remote or relocation)

Real products, generated worlds.

I'm Dionisis, an AI Designer with 5 years at Superside producing hero, lifestyle and ad imagery for Pfizer, Amazon, Intuit, QuickBooks and Lucid. I put real products into fully generated scenes and keep every detail accurate.

Scroll to work ↓
PfizerAmazonIntuitQuickBooksLucidMarqeta PfizerAmazonIntuitQuickBooksLucidMarqeta

(01) Selected AI work

Three image systems · 2026

Hush bedside light glowing on an oak nightstand
01

Hush

One product. Every surface. Same standard.

A repeatable AI pipeline for a bedside light for tinnitus: 12 shots, 3 layouts and 3 ad formats, with the product locked from a single packshot.

  • Product-in-scene
  • Prompt kit
  • QA scorecard
  • Photoshop finish
View case study →
Norra modular sofa in a sunlit living room
02

Norra

Built in modules. Made for staying.

A reference-locked workflow for a modular sofa. Arm radius, welt seam and oak legs match the real product in every frame. 9 of 11 generations approved.

  • Reference library
  • Ecommerce
  • Paid social
  • Responsive hero
View case study →
Solenne hybrid mattress in a Japandi bedroom
03

Solenne

One mattress. Every room.

One generated master packshot taken across seven interiors from 2,700K to 5,600K, plus lifestyle and ad formats. Same band, piping and label every time.

  • Product DNA
  • Scene templates
  • Label QA
  • Landing page
View case study →

(02) About

I make AI imagery that's ready for production. I specialise in photoreal product and lifestyle work: real products placed into generated scenes, then composited and retouched to a final standard. I also design the pages they live on, so I judge every image in a live layout, not as a standalone render.

0Years at Superside
0Years in design
0Image systems shown here
0Approved images across them
01AI generation

Hero, lifestyle and advertising imagery made with GPT Image, Gemini / Nano Banana, Midjourney, Krea and Weavy, using reference-image prompting, image-to-image and reusable prompt templates.

02Product fidelity

The product stays accurate in every scene. I lock silhouette, materials and branding from a reference, then fix drift with inpainting, outpainting and iterative editing instead of rerolling.

03Compositing & retouching

Generation gets an image close; Photoshop finishes it. I composite, retouch and colour-match outputs into one consistent, production-ready set.

04UI/UX & layout

Web, ecommerce and landing-page layouts in Figma, built inside client design systems. Every image is briefed for its placement: copy space, safe zones, thumbnail legibility.

(03) Experience

  1. 2021 — 2026 AI & Graphic Designer Superside AI hero, lifestyle and ad imagery for enterprise and startup clients, plus web and landing-page design in Figma.
  2. 2018 — 2019 Graphic Designer Codex In-house graphic design.
  3. Since 2015 Freelance Graphic Designer 99designs Branding, marketing and digital work for international clients, won through competitive briefs.

Toolkit

  • GPT Image
  • Gemini / Nano Banana
  • Midjourney
  • Krea
  • Weavy
  • Photoshop
  • Figma
  • Illustrator
  • After Effects
  • Adobe XD
  • Learning: Blender

Education

  • IEK DeltaAnimation, Interactive Technology, Video Graphics & Special Effects · 2019–2021
  • University of IoanninaBachelor's degree, Economics · 2014–2020

Languages

Greek (native) · English (professional)