Launching the MVP of PickifyAI: Using LLMs to solve the "Gift-Finding Fatigue"

Today is a big day! I’m officially sharing the MVP of PickifyAI.

The Problem: We’ve all been there—scrolling through endless gift guides that feel generic and outdated.

The Solution: I built PickifyAI to turn the gift-finding process into a 30-second conversation. By using the Claude API for structured reasoning, the app understands the recipient's personality rather than just matching keywords.

What's under the hood?

• Frontend: React 19 (keeping it fast and modern).

• AI: Claude-3.5-Sonnet for high-intent recommendations.

• Database: Neon DB.

The "GEO" Experiment: One thing I’m particularly excited about is optimizing the site for Generative Engines. I’m testing how search-augmented LLMs perceive the product data to ensure we’re discoverable in the new age of AI-driven search.

I’d love for this community of builders to try it out and give me some brutal feedback on the flow!

Check it out here: https://pickifyai.com/