Launching today
Runner AI
Build, optimize, and scale your AI-native store
442 followers
Build, optimize, and scale your AI-native store
442 followers
Don't just prompt a website; prompt revenue. Runner AI builds AND optimizes. It continuously runs experiments in the background, automatically turning visitors into buyers.
Runner AI
Hey Product Hunt!
I'm Weizhi, founder of Runner AI. Excited to share what we've been building over the past 5 months: an AI-native commerce stack that autonomously builds, tests, and optimizes your store.
Before starting Runner AI, I spent five years at Google and DeepMind — building data infra and Gemini. That shaped how I think about products: when data is unified and AI can see the full picture, it doesn't just help — it proposes changes, learns from real user behavior, and gets smarter with every iteration.
But today, traditional e-commerce doesn't work that way. They stitch together 20+ tools that don't talk to each other. Your email tool doesn't know what your checkout learned. Your analytics can't inform your pricing. Because nothing connects, nothing compounds. You're running experiments in silos while customer behavior shifts daily.
We built Runner to bring that self-optimizing logic to commerce, grounded in the belief that the best way to grow is to experiment, learn, and iterate continuously.
Most e-commerce platforms are built for humans to operate, with AI added later as a feature. We designed it differently — a commerce stack built from the ground up for AI to observe, act, and improve your conversion rates automatically.
Runner puts everything under one roof: storefront, analytics, experiments, pricing, promotions, user behavior — all unified, all visible to AI. When checkout improves, recommendations get smarter. When pricing learns, promotions adapt.
How it works:
Describe & Build: Describe your vision → AI builds everything (design, copy, functionality).
Observe: AI sees everything: drop-offs, abandoned products, winning copy, and pricing friction.
Optimize: AI acts on everything: rewrites headlines, adjusts layouts, optimizes flows, and tests pricing.
Scale: Winners get scaled. Losers get replaced. 24/7.
What makes us different:
No themes, no templates: AI architects a custom store from scratch.
No "App Tax": Core tools are built-in, not paid plugins.
Full transparency & control: You see every experiment, set boundaries, and adjust anytime.
This is v1. We'd love for you to try it and tell us what you think.
→ Try Runner AI free
Weizhi Li
Founder & CEO, Runner AI
The "continuous experiments in the background" angle is really compelling — most A/B testing tools require you to manually set up experiments which nobody actually does consistently.
Curious how it handles low traffic sites where statistical significance takes forever to reach? That's always been the frustration with traditional CRO tools for early-stage products.
Congrats on the launch!
@n8nship Great question! For low-traffic sites, we introduce our AI creative studio which generates a range of high-converting product assets while our AI ads system creates a closed-loop. This combination lets us gather statistically significant insights faster, so you can start optimizing your site's performance even in the early stages without waiting for massive traffic volumes.
Runner AI
@n8nship You are right. low traffic is where traditional A/B testing breaks down. Our AI uses multi-armed bandit approaches that adapt in real time, so even with lower traffic, the system learns and shifts toward better-performing variants faster instead of waiting for a fixed sample size. On top of that, we also have built-in marketing integrations — ads, social media, and more — to help drive traffic to your store in the first place.
Prava
Awesome!
Are you helping stores with product feed optimization for agentic commerce?
@kaiserrr Absolutely! Runner AI continuously detects friction and autonomously A/B tests to optimize the product feed. Because our entire architecture is AI-native from the ground up, your product data is perfectly structured for agentic commerce and AI shopping bots by default.
Prava
@stacy_liu2 Awesome!
We should chat after the launch. Atb!
Runner AI
@stacy_liu2 @kaiserrr yes. let's chat
Runner AI
@kaiserrr Absolutely! Product feed optimization is built right in. We handle formatting across OpenAI ACP and Google UCP — so your store is ready for agentic commerce out of the box.
The "self-optimizing" framing is what caught my attention here. Most ecommerce AI tools I've seen are basically just suggestions you still have to manually implement. Having the system actually run experiments and adjust on its own is a totally different level.
One question though, how much control do store owners have over what gets tested? Like can you set guardrails so it doesn't experiment with things you'd rather keep fixed (brand elements, pricing floors, etc)?
Runner AI
@mihir_kanzariya Great question! You define the boundaries — pricing, visuals, ads, social media, promotions, etc. You also control the traffic split, so you decide how much of your audience sees a test before it rolls out.
From an agency perspective, this could accelerate client onboarding significantly. Interested to see the template variety.
@sandy_liusy As an agency, you’ll love that we can support multi-store management-you can handle multiple clients’ sites all in one place. As a multi-brand owner, you can easily create tailored strategies for each unique product line, drive growth across your entire product range.
@sandy_liusy Absolutely! For agencies, this takes the heavy lifting out of the 'zero to one' phase. The coolest part? There are zero templates. You simply prompt the AI with the specific style, vibe, and branding instructions your client needs, and it custom-codes the site natively. It means you can deliver fully custom-tailored stores in a fraction of the time.
How does it decide which experiments to run first — is that AI-driven or do you set the priorities manually?
@jozsef_orsos Great question! Yes, it is 100% AI-driven! Runner AI acts like an autonomous growth engineer. It actively analyzes real user behavior—every click, scroll, and bounce on your site. Once it detects friction points or high-impact opportunities, the AI automatically prioritizes and launches the tests for you.
App Finder
Looks like a really innovative product. Has a "real" store already been built with it?
Runner AI
@konrad_sx Thanks for asking! Yes, we've been working with a handful of design partners who have live stores on the platform with real traffic. Still early days, but the results have been really encouraging.
App Finder
@weizhi great to hear that, wish you success!
Runner AI
@konrad_sx Thank you!