DEV Community

Play Button Pause Button
ANIRUDDHA  ADAK
ANIRUDDHA ADAK Subscriber

Posted on

AI LinkedIn Profile Generator

This is a submission for the Google AI Studio Multimodal Challenge

What I Built

I'm excited to share my project, the AI LinkedIn Profile Generator, a comprehensive tool designed to help professionals craft a polished and cohesive online brand in minutes. In today's digital-first world, a compelling LinkedIn profile is non-negotiable, but creating one from scratch—with professional photos, a custom banner, and sharp, engaging copy—can be a significant hurdle.

My applet streamlines this entire process. By providing simple inputs like their profession, industry, and a desired aesthetic, users receive a complete branding package generated by AI:

  • 10 unique, professional profile photo options.
  • A custom-designed, abstract banner image that fits their industry.
  • A powerful, attention-grabbing headline.
  • A well-articulated professional summary for their 'About' section.

But the real magic is in the personalization. Users can select any of the AI-generated photos and enter an editing modal to refine them further with simple text prompts. This ensures the final output isn't just AI-generated, but personally curated.

Demo

Check out the deployed applet here: Link

Here's a glimpse into the user experience:

1. The intuitive input form where users define their professional brand:

2. The complete set of generated assets—banner, headline, summary, and a gallery of photo options:

3. The powerful in-app image editor, allowing for precise, text-based modifications:

(As this project relies on the latest models, I've also prepared a short video walkthrough to ensure the full functionality can be reviewed.)

See video here on x/twitter:

How I Used Google AI Studio

Google AI Studio and the Gemini API are the engine driving this entire application. I strategically employed a suite of models to handle the diverse multimodal requirements:

  • imagen-4.0-generate-001: This state-of-the-art text-to-image model was my choice for generating the high-quality 1:1 profile headshots and the 16:9 banner image. Its nuanced understanding of style and professional aesthetics was critical to producing relevant and impressive visuals.
  • gemini-2.5-flash: For generating the headline and summary, I turned to gemini-2.5-flash. The standout feature for me was the ability to enforce a responseSchema. This allowed me to request and receive perfectly structured JSON every time, which eliminated the need for fragile string parsing and made integrating the text into the UI incredibly robust.
  • gemini-2.5-flash-image-preview: This model is the star of the show, powering the interactive editing feature. It accepts both an image and a text prompt as input to generate a new, modified image. This is what allows a user to ask the AI to "add glasses" or "change the background to a blurred office," making the app truly interactive.

Throughout the development, AI Studio served as my command center for prompt engineering, allowing me to rapidly test and iterate on prompts to achieve the perfect balance of creativity and professionalism in the generated assets.

Multimodal Features

This applet is fundamentally a multimodal experience, creating a seamless and powerful workflow that blends user input, text, and images.

  1. Text-to-Image Generation: At its core, the app translates a user's professional identity (described in text) into a rich set of visual assets (images). This is the first step in bringing their brand to life visually.

  2. Image-and-Text-to-Image Editing: This is the most innovative feature and where the multimodal capabilities truly shine. Instead of simply accepting the initial AI output, users can enter a creative dialogue with the model. They select an image they like and use natural language to direct changes. This multimodal interaction dramatically enhances the user experience by:

    • Granting Agency and Personalization: It puts the user in the director's chair. They can fine-tune the AI's creation to perfectly match their personal style and vision, moving beyond a one-shot generation.
    • Enabling an Iterative Creative Process: It encourages experimentation. A user can try multiple prompts on the same image, creating an engaging and fun feedback loop that ultimately leads to a more satisfying and personalized result.

By weaving these multimodal features together, the AI LinkedIn Profile Generator elevates itself from a simple generator to an interactive branding partner.

Top comments (0)