What I build

Software for e-commerce

An AI analyst for your store

Ask about revenue, conversion, products or returns in plain English and get the answer with the numbers behind it.

Operations dashboards

Orders, returns, stock and suppliers in one view your team actually uses, updated automatically.

Wholesale and B2B ordering

A portal where trade customers see their prices and reorder in a few clicks, instead of emailing PDF order forms.

Workflow automations

Returns, refunds, supplier reorders and alerts handled by rules, with people stepping in only where judgement is needed.

Product content tools

AI drafts product descriptions and SEO copy from your specs, and a person approves every word before it's published.

Reliable data pipelines

Millions of product and order records imported, cleaned and kept in sync, with retries and error logs.

AI, right now

Where AI genuinely helps

Most e-commerce teams have more data than they can use: product, order, marketing and support data spread across Shopify, ad platforms and spreadsheets. The questions are simple (“what should we restock?”, “why are returns up?”); getting answers takes someone hours.

AI makes it practical to ask those questions directly, as long as the answers are grounded in your real numbers. That’s how StoreFilter works: 2.5 million records, store matching with vector search, and GPT-4 answers that show their working.

  • Questions instead of reports

    Founders ask "which products have rising returns this month?" and get an answer, instead of waiting for someone to build a report.

  • Content at catalogue scale

    Descriptions, alt text and SEO copy drafted from structured product data, then reviewed by your team.

  • Support answers from your policies

    Shipping, returns and sizing questions answered from your own policies, with sources, and a hand-off to a person when needed.

  • Early warnings

    Automatic alerts when returns spike, stock runs low or a best-seller suddenly stops selling.

Proof

A real project in this industry

StoreFilter chat answering a question about a Shopify store's conversion rate, with the calculation, insights and follow-up questions

RAG and AI chatbots

StoreFilter: an AI analyst for any Shopify store, built on 2.5 million records

An AI analytics chat for e-commerce, built with Next.js, Claude Code, Supabase, pgvector and GPT-4. Ask about any Shopify store, get answers with charts.

Client
StoreFilter
Role
Full-stack developer
When
2025
  • Next.js
  • Claude Code
  • Supabase
  • PostgreSQL + pgvector
  • Edge Functions
  • OpenAI GPT-4

Data, security and access

  • Store data is read with the narrowest access that does the job, read-only wherever possible.
  • Customer personal data is kept to what's needed and only visible to the roles that need it.
  • AI provider keys stay on the server, with spending limits.
  • AI answers are calculated from your real data and show the figures and formula, so they can be checked.

How a project runs

  1. Call

    A 30-minute call about what you need, who it is for, and what success looks like.

  2. Scope

    A written scope: features, architecture, timeline, price range and risks.

  3. Build

    Working increments you can click through, not a big reveal at the end.

  4. Ship

    Live on your domain and hosting, with a handover so you know how it runs.

More in how I work, or see rag and ai chatbots.

Questions

Common questions

Shopify app or custom software?

If an existing app solves the problem, use it. Custom software is worth it when your workflow is specific to you, when you're paying for several apps that don't talk to each other, or when the data you need lives in more than one system.

Will an AI analyst make up numbers?

It shouldn't, and it's built so it can't quietly. The numbers are pulled from your data first and the AI explains them, showing the inputs and the calculation, as in the StoreFilter answers.

Does it work with Shopify and other platforms?

Yes. Shopify, payment providers and marketing tools all offer APIs and exports. Which ones we connect is decided in the scope, based on where your data lives.

How much does this cost?

AI and data projects are quoted at a fixed price after a 30-minute call. Smaller jobs can run hourly at $30/hour.

Want answers from your store data, not more reports?

A 30-minute call is enough to tell whether it fits, what the first version should include, and how long it takes.

Book a 30-minute call