AM Jarvis is the second product case study we publish, and the one closest to the daily work of our ecommerce clients. Mustafa co-builds it as a SaaS suite for Shopify and WooCommerce store owners, and it grew out of the jobs we kept seeing stores do by hand: importing products one at a time, fighting Merchant Center warnings, and editing hundreds of listings through spreadsheets.
What AM Jarvis Does
AM Jarvis is a live product at jarvisbyam.com, with public signup and the app itself at app.jarvisbyam.com. It brings 14 Pro features into one dashboard, each documented on its own feature page. The core store operations tools are:
- A bulk product importer that pulls products from Shopify, WooCommerce, AliExpress, Amazon and eBay, with AI rewriting of titles and descriptions.
- A Google Merchant Center violation scanner and fixer.
- Bulk product editing with a preview before changes apply and an undo after they do.
- A site builder that sets up an LLMS silo structure.
- Niche research, a lead finder and a rank tracker.
- Programmatic SEO with a uniqueness gate.
The other Pro features cover a Chrome extension for importing from AliExpress and Amazon, a dropshipping suite, Social Autopilot for Pinterest, Instagram and Facebook, YouTube Studio, Viral Finder and Marketplace Research for Amazon, Etsy and eBay pricing. The keyword and niche research tool covers 18 countries, and the rank tracker includes local map positions.
Alongside the paid features, the site offers 73 free tools across 11 categories, including Shopify, PDF, image, YouTube, SEO and pricing calculators. They need no signup, and most run entirely in the browser, so files never leave the user's device.
It runs on the user's own AI API keys with no markup on AI usage. Supported providers are Anthropic (Claude), OpenAI, Google Gemini, DeepSeek, OpenRouter and Groq.
Store Problems AM Jarvis Was Built to Solve
Store owners on Shopify and WooCommerce run into the same four problems again and again, and each one eats time that should go into marketing and customers.
Slow product imports
Adding products from a supplier or marketplace usually means copying titles, descriptions, images and variants by hand, or wrestling a CSV into the right format. Copied supplier text is also the same text on every other store selling that product, which does nothing for search visibility or conversion.
Merchant Center suspensions
Google Merchant Center can disapprove products or suspend a whole account over policy problems, such as prices that do not match the landing page, missing product identifiers or policy pages that are hard to find. For a store that depends on Shopping ads, a suspension stops a major sales channel, and the warnings do not always make the fix obvious.
Risky bulk edits
Changing hundreds of product titles, tags or prices at once is where stores break things. A wrong column in a spreadsheet import can overwrite good data across the catalog, and platform bulk editors rarely show a clear before-and-after or offer a way back.
Research spread across too many tools
Niche research, rank tracking and lead finding each usually mean a separate subscription and a separate login. Small teams end up paying for several tools and using a fraction of each.
How We Built AM Jarvis
- A bulk importer with AI rewriting. Products come in from Shopify, WooCommerce, AliExpress, Amazon and eBay in bulk, and AI rewrites titles and descriptions so the store is not publishing the same supplier copy as everyone else. The store owner stays in control of what is imported and published.
- A Merchant Center violation scanner and fixer. The scanner checks a store for the kinds of problems that lead to Merchant Center violations, and the fixer helps apply corrections so products can be resubmitted. The goal is to turn a vague warning into a specific list of changes.
- Bulk editing with preview and undo. Every bulk edit shows a preview of exactly what will change before anything is written to the store, and applied changes can be undone. This was a deliberate design choice: speed only helps if a mistake is cheap to reverse.
- Bring-your-own AI keys. Instead of reselling AI usage, AM Jarvis runs on the user's own API keys for Anthropic (Claude), OpenAI, Google Gemini, DeepSeek, OpenRouter or Groq, with no markup. Usage is billed by the provider directly, at the provider's own rates, and the user chooses which model handles which job.
- A site builder with an LLMS silo structure. New sites are set up with the same parent, child and sibling page structure we use on client sites and on Boiler Decoder, so internal linking and topical coverage are planned from the start.
- Programmatic SEO with a uniqueness gate. Generating pages at scale is easy. Generating pages that are different enough to deserve indexing is the hard part. The uniqueness gate checks generated pages before they publish, so near-duplicates are held back instead of flooding the site with thin content.
- Research and tracking in the same place. Niche research, a lead finder and a rank tracker sit alongside the store tools, so the research that shapes a store and the tracking that measures it share one login.
AM Jarvis Results and Lessons So Far
We do not publish user counts, revenue or growth figures for AM Jarvis. This case study is about the problems it addresses and the way it is built, which is what matters when you are deciding whether we can build something similar for you.
The clearest outcome is what building it has taught us. Designing preview and undo into bulk operations from the first version is far easier than adding it later. Letting users bring their own AI keys removes a pricing problem and a trust problem at once, because the user can see exactly what the AI costs. And a uniqueness check belongs in any system that generates pages, because the easiest way to damage a site's search visibility is to publish thousands of pages that say the same thing.
What AM Jarvis Means for Client Work
AM Jarvis shows the kind of product and marketing thinking we bring to client work.
- SaaS builds from people who run one. If you are planning a subscription product, our SaaS development work is shaped by the same decisions we make on AM Jarvis: billing models, user trust, safe bulk operations and integrations with third-party platforms.
- Merchant Center experience. Clients running Shopping campaigns benefit from the time we have spent on feed quality and policy problems. It feeds directly into our Google Shopping ads management.
- Practical AI automation. The bring-your-own-key model and human review points are the same principles we use in AI automation projects: the client controls the provider and the cost, and people approve the changes that matter.
- Programmatic SEO with guardrails. The uniqueness gate reflects how we run programmatic SEO for clients: templates and data are only worth publishing when each page answers a distinct search intent.
If you have a store operations problem that a custom tool could solve, or a SaaS idea you want to scope properly, get in touch. You can also read how we built our content site in the Boiler Decoder case study.