← sources

Building a 7-module AI website audit engine. What would you use for the backend? Day 2.5

source: reddit (reddit:AI_Agents)published 2026-10-23 04:31ingested 2026-10-24 02:35

retained text

Quick clarification before I get into this: I'm not an AI bot, despite what a few Reddit users seem to think. 😂

I'm a UX/UI designer building an AI product using AI tools. I also use AI to proofread my posts because English isn't my first language. But there's a very real, slightly overwhelmed human behind this experiment.

I'm working with a $1,500 budget and very little hands-on backend experience.

The frontend is being built in Lovable, but I want my website audit engine to operate independently.

The proposed architecture:

1. User submits a URL.
2. Backend retrieves the webpage and collects evidence.
3. Seven specialized AI modules analyze UX, UI, copywriting, content, accessibility, SEO, and marketing/conversion.
4. Results are validated, deduplicated, and prioritized.
5. A structured report is returned to the frontend through an API.

I've finished drafting the seven analysis prompts, but I haven't built the backend yet.

My priority is to create a working prototype and test the accuracy and consistency of the results before investing heavily in infrastructure.

I'm not looking to build a complicated multi-agent system just for the sake of it. If a simpler architecture works, I'd prefer that.

**For anyone who's built something similar: What platform or stack would you recommend for a beginner building this kind of pipeline?**

I'm particularly interested in keeping costs predictable while running repeated tests.

document fetched 2026-10-24 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1xd4d9h/building_a_7module_ai_website_audit_engine_what