Collect on a schedule
One shared job reads products, stock and orders from the source systems. It finishes the collection and checks the records before releasing an update.
AI practice · Engineering & delivery
AI implementation
Useful tools. Thoughtful websites.
My starting point is the business problem. I turn my experience in e-commerce, marketing and product operations into clear requirements, shape the front-end experience, and guide AI-assisted development through testing and release.
01
Yola Workbench
Better analysis.
Stock files people can use.
I wanted to close the gap between the ERP’s standard reports and the questions our team needed to answer—not replace the ERP.
Example task
Build the view the team needs.
BrandChannelDate range
Choose products and export columnsBuyer-ready file
Two everyday jobs.
One small desktop tool.
The ERP’s reporting dimensions were too limited, and colleagues could not easily export current stock in the format each buyer needed. I designed a workbench that brings flexible analysis and customized inventory exports together.
One shared update.
Ready for every colleague.
The cloud handles collection; the desktop handles everyday analysis and exports.
Zoho + marketplace recordsProducts · stock · orders
Keep a small index
A lightweight database tracks record identities, versions and job status, so the tool knows which update is ready.
Store the larger files
Checked data bundles and daily stock snapshots live here. The desktop downloads what it needs for analysis and buyer files.
Yola WorkbenchCheck the update, then refresh the local data.
Easier to maintain
The collection job and its status are managed in one place, instead of running separately on every colleague’s laptop.
Less storage to pay for
Reuse unchanged history, keep useful daily stock snapshots, and expire temporary processing files after seven days. That limits repeated storage and downloads.
My contribution
I defined the filters, export formats and day-to-day tasks, shaped the interface, and reviewed results with colleagues. AI assisted with implementation and debugging, including the data connections behind the screens.
Keeping the information useful.
Updates need to stay practical: bring in the changes, retain useful history and keep a working copy if a refresh fails.
Read current recordsCheck the update
Changed productsBring in the changes
Unchanged recordsReuse existing data
Refresh the data.
Keep the work simple.
APIs connect the tool to the existing systems instead of relying on manually collected spreadsheets. After the first download, changed records can update the local copy without transferring the same history again.
Latest downloadIncomplete
Previous checked dataStill available
Next actionRetry the update
Keep a usable copy.
An incomplete update should not erase the data already on the desktop. The previous checked copy remains available while the update is retried; it is not presented as a successful new refresh.
02
Website development
From an operating need
to a website people can use.
My role combines requirements, front-end direction and hands-on review. AI helps produce and revise the code; I guide what gets built and whether it works for the business.
SkinNote · Shopify
A retail storefront for product discovery and clearance sales. Theme development lets me shape merchandising, content and the shopping experience.
Yola · Zoho Commerce
A wholesale inquiry and ordering site connected to the ERP. Buyers need clear product information and a dependable path from selection to inquiry.
UserWholesale buyer
NeedFind stock and request an order
RequirementsProduct details · stock · inquiry
ReviewCan a buyer complete the task?
Start with a usable brief.
I provide the documentation, operating needs, audience and examples. Together, we turn those into specific pages, functions and review criteria before coding.
STORESearch
Clear offer. Clear next step.
Illustrative layout
Get the structure right first.
AI helps build the initial layout. I review the customer journey, content and priorities, then ask for changes before we invest in visual polish.
ModulesProduct cards · filters · cart
ContentProduct information & SEO
VisualsType · spacing · image treatment
My reviewBusiness fit and usability
Customize the modules.
Then refine the visuals.
I guide the front-end details: product presentation, page modules, interactions, mobile layouts and visual consistency. My commerce and SEO experience informs those decisions.
Recover the existing theme
shopify theme pullBuild and preview changesshopify theme devCheck and upload a review versionshopify theme checkshopify theme push --unpublishedMove the draft into its platform.
For Shopify, we work with theme files and CLI previews. For Zoho Commerce, we adapt the platform’s templates and native shopping behavior. A local demo is checked again in its actual platform.
Phone & desktopLayout and navigation
Product selectionCorrect option and quantity
Cart or inquiryComplete the customer journey
After a fixRepeat the affected checks
Test a complete customer task.
I review the preview at both screen sizes and walk through the important actions. Problems return to development; the corrected flow is tested again before release.
Before releaseKeep the previous version
ApprovalReview the actual preview
ReleaseMove approved changes live
After releaseCheck the live result again
Publish the reviewed version.
After approval, the reviewed changes go live. We retain the previous version, check the live site and handle follow-up fixes through the same review cycle.
A small example of the review.
For a wholesale cart, a good-looking button is not enough: the selected product and quantity must reach the cart once, and the buyer needs a clear result. The public code example isolates that behavior so it can be tested.
View the cart example on GitHub