Annalytical Design | Anna Eunbyeol's Portfolio

Emoji Day

the goal was to drive app visits (new & returning) around July 17 and increase in-chat emoji use as a lightweight way to connect across language barriers, tying the event straight back to the product's core value. I led creative across the funnel: App Store motion banner (built in Figma Motion) and in-app popup (IAC) assets, designed against a mapped user journey (App Store discovery → in-app awareness → in-chat action). (No splash screen at this stage. The post-splash surface came next, with Pet Day.) Copy tested across 4 variants and localized for six markets (US, GB, DE, FR, IT, KR).

To bring the "Judgmental Cat" concept to life within a tight timeframe, I developed a highly efficient, AI-collaborative production pipeline, acting as both the creative director and technical executor.

Reverse Engineering Visuals with AI: I began by sourcing specific visual references that matched the desired mood. Instead of guessing prompts, I used AI to analyze these references and extract precise, descriptive text prompts. I then fed these exact prompts into Gemini to generate the core still images.

Keyframing AI Video Sequences: To create fluid motion, I didn't rely on random video generation. I carefully generated specific first and last frame images using Gemini, and then used AI video tools to bridge these frames into seamless, controlled video sequences based on my storyboard.

The "Kim Hamster" Editing Technique: To capture the chaotic, viral energy of the Korean "Kim Hamster" meme, I employed rapid fire jump cuts. This fast-paced editing not only kept the viewer's attention but also strategically masked the typical "AI generated" feel, making the video look like a raw, native social media clip rather than a synthetic render.

Voice Acting & Modulation: To give the cat its signature sassy persona, I recorded the voiceover myself and applied voice modulation to achieve the perfect comedic pitch.

Pet Day (Cat & Dog)

I introduced the post-splash surface here (reused for later events), with art built on a "face + pet" concept to invite users to update their profile photos. Rolled out with permarket cultural adaptation: Cat Day ran across nearly all markets; Dog Day deliberately excluded Muslim majority markets (plus one market already running another campaign), because dog imagery is culturally sensitive there.

To bring the "Judgmental Cat" concept to life within a tight timeframe, I developed a highly efficient, AI-collaborative production pipeline, acting as both the creative director and technical executor.

Reverse Engineering Visuals with AI: I began by sourcing specific visual references that matched the desired mood. Instead of guessing prompts, I used AI to analyze these references and extract precise, descriptive text prompts. I then fed these exact prompts into Gemini to generate the core still images.

Keyframing AI Video Sequences: To create fluid motion, I didn't rely on random video generation. I carefully generated specific first and last frame images using Gemini, and then used AI video tools to bridge these frames into seamless, controlled video sequences based on my storyboard.

The "Kim Hamster" Editing Technique: To capture the chaotic, viral energy of the Korean "Kim Hamster" meme, I employed rapid fire jump cuts. This fast-paced editing not only kept the viewer's attention but also strategically masked the typical "AI generated" feel, making the video look like a raw, native social media clip rather than a synthetic render.

Voice Acting & Modulation: To give the cat its signature sassy persona, I recorded the voiceover myself and applied voice modulation to achieve the perfect comedic pitch.

World Tourism Day

invited users to set a favorite travel photo as their profile ("show where you've roamed") and meet friends from countries they want to visit. Ran across all markets for ~18 days around Sept 27 (timed to catch three weekends plus a Korean holiday). Assets: four post-splash variants + a store event card produced in three ethnicities (white / Asian / Black) for inclusive representation. Results still in review.

To bring the "Judgmental Cat" concept to life within a tight timeframe, I developed a highly efficient, AI-collaborative production pipeline, acting as both the creative director and technical executor.

Reverse Engineering Visuals with AI: I began by sourcing specific visual references that matched the desired mood. Instead of guessing prompts, I used AI to analyze these references and extract precise, descriptive text prompts. I then fed these exact prompts into Gemini to generate the core still images.

Keyframing AI Video Sequences: To create fluid motion, I didn't rely on random video generation. I carefully generated specific first and last frame images using Gemini, and then used AI video tools to bridge these frames into seamless, controlled video sequences based on my storyboard.

The "Kim Hamster" Editing Technique: To capture the chaotic, viral energy of the Korean "Kim Hamster" meme, I employed rapid fire jump cuts. This fast-paced editing not only kept the viewer's attention but also strategically masked the typical "AI generated" feel, making the video look like a raw, native social media clip rather than a synthetic render.

Voice Acting & Modulation: To give the cat its signature sassy persona, I recorded the voiceover myself and applied voice modulation to achieve the perfect comedic pitch.

AI-generated assets: the imagery and video across these events were produced with Nano Banana (Gemini) and Veo, then composed and animated in Figma Motion. The same AI-integrated pipeline as the perf-marketing creative.

Consistency: everything built in Figma Motion so the seasonal creative matches the look and cadence of the perf-marketing pipeline, not a separate one off style.

To bring the "Judgmental Cat" concept to life within a tight timeframe, I developed a highly efficient, AI-collaborative production pipeline, acting as both the creative director and technical executor.

Reverse Engineering Visuals with AI: I began by sourcing specific visual references that matched the desired mood. Instead of guessing prompts, I used AI to analyze these references and extract precise, descriptive text prompts. I then fed these exact prompts into Gemini to generate the core still images.

Keyframing AI Video Sequences: To create fluid motion, I didn't rely on random video generation. I carefully generated specific first and last frame images using Gemini, and then used AI video tools to bridge these frames into seamless, controlled video sequences based on my storyboard.

The "Kim Hamster" Editing Technique: To capture the chaotic, viral energy of the Korean "Kim Hamster" meme, I employed rapid fire jump cuts. This fast-paced editing not only kept the viewer's attention but also strategically masked the typical "AI generated" feel, making the video look like a raw, native social media clip rather than a synthetic render.

Voice Acting & Modulation: To give the cat its signature sassy persona, I recorded the voiceover myself and applied voice modulation to achieve the perfect comedic pitch.

Azar In-App Events (Seasonal ASO)

How do you keep a seasonal event from being a gimmick and make it feel like the product itself?

Other Works

Other Works