xuy

xuy / sgo

Public

Optimize any entity against evaluator populations using LLMs and counterfactual probes as gradient estimators

16
1
100% credibility
Found Mar 27, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

SGO simulates diverse groups of realistic people evaluating products, resumes, pitches, or content using AI, providing scores, segment insights, and prioritized improvement suggestions.

How It Works

1
💡 Discover SGO

You hear about SGO when you want quick, honest feedback on your product idea, resume, or pitch from a realistic crowd of people.

2
📝 Describe Your Idea

You type in a description of what you're working on, like your landing page or cover letter, so everyone understands it clearly.

3
Pick Your Crowd
🇺🇸
Everyday Americans

Use a big list of real-life people from all walks of life to get broad opinions.

🎯
Target Experts

Build a special group like startup bosses or hiring managers tailored to your goal.

4
⭐ See the Scores

Watch as each person in the group rates your idea from 1 to 10 and shares what they like or worry about.

5
📈 Get Smart Suggestions

Discover the top changes that would boost your scores the most, like adding a free option or trust badges, ranked by impact.

6
🔄 Try a Change

Update your idea with the best suggestion and run the check again to see the improvement.

🎉 Win Better Feedback

Your idea now scores higher across the group, ready to impress real people with confidence.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is sgo?

SGO lets you optimize anything—a landing page, resume, pitch deck, or even GitHub profile—by simulating evaluations from 50 realistic personas drawn from a 1M-person US census dataset. Paste your entity, pick a target audience, and it scores reception across segments like solo devs vs. enterprise CTOs, then computes a "semantic gradient": prioritized changes (e.g., "add free tier +1.8 points") via LLM counterfactual probes. Built in Python with OpenAI-compatible APIs, a FastAPI web UI, Docker support, and CLI for quick runs—all for pennies in minutes.

Why is it gaining traction?

Unlike vague LLM feedback or slow surveys, SGO delivers census-grounded panels that avoid archetype collapse, plus bias audits matching human cognition for credible results. The gradient ranks fixes by impact on the "persuadable middle," helping optimize GitHub Actions workflows, Steam games, or PC setups without A/B traffic. Devs dig the instant iteration loop: tweak, re-eval same cohort, track deltas.

Who should use this?

SaaS founders optimizing landing pages or pricing for buyer personas; job hunters tuning resumes against hiring managers; indie devs refining GitHub profiles or game topologies; indie hackers iterating pitches for VCs. Great for anyone needing fast synthetic user testing before real launches, like tweaking Minecraft servers or Google Maps integrations.

Verdict

Try it for hypothesis generation on optimize-anything tasks—solid docs, intuitive web app, and MIT license make early experiments low-risk despite 16 stars and 1.0% credibility score. Still maturing (no tests visible), so validate gradients with real users; pair with scikit-optimize for production.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.