Solution · E-commerce
Supplier list → platform-ready catalog
Drop in a vendor PDF or photo. Get a Shopify, WooCommerce, or Square CSV that imports cleanly — currency split, units normalised, variants resolved.
Requirement confirmed
Fast delivery
Continuous security
How adoption works
Three phases, each ending in something you can sign off
An adoption project fails on scope, not on models. Each phase below closes with a written artefact, so the next one starts from an agreement rather than an assumption.
Requirement confirmed
We inventory the documents and systems already in use, agree which fields matter and how each is read, and draw the line where a person reviews rather than the workflow decides.
Output of this phaseA written scope note naming every field, every reading rule, and every human checkpoint.
Fast delivery
The workflow is assembled from generic platform tools rather than built for one customer. Real sample documents run first, and volume follows once the sample set is clean.
Output of this phaseA running workflow on real sample documents, with the cost per run measured before scale-up.
Continuous security
Every run leaves a tool-invocation record. Retention windows, deletion behaviour and the sub-processor list are published pages that a security reviewer can check without asking us.
Output of this phaseAn audit trail per run, and a documented retention, deletion and sub-processor position.
Scope of the exchange
What you hand over, and what comes back
A catalogue engagement is scoped by the source formats you receive and the importer you have to satisfy. Both are named here before any work starts.
- Your organisation providesSupplier price lists, spec sheets and product photos, as PDFs, spreadsheets or phone pictures of a printed catalogue.Convilyn deliversEvery product read out of them, with each field carrying the page it came from.
- Your organisation providesThe importer template your marketplace or storefront actually accepts, exported from your own seller centre.Convilyn deliversThat template's real columns read and filled in place, so no marketplace layout is assumed or hardcoded.
- Your organisation providesCurrency handling, tax-inclusive or exclusive pricing, unit conventions, and how variants map to your option names.Convilyn deliversCurrency split into its own column, units normalised, and variants resolved against your option scheme rather than invented.
- Your organisation providesProduct images and any naming rule the importer expects for them.Convilyn deliversImages matched to the right product row and named to that rule, with unmatched images listed rather than dropped.
- Your organisation providesThe person who approves pricing before anything is listed.Convilyn deliversA tool-invocation record per run. Prices are transcribed, never recalculated, and every override is logged.
From supplier list to live catalog
Same job, two routes. Manual is seven steps of column hunting and importer retries; the AI route folds the platform-specific schema into the upload itself.
Manual workflow07 steps
Receive supplier PDF or photo
Open Excel, type each row by hand
SKU, name, variants, currency, MOQ — re-key every field, every page.
Look up target platform CSV schema(this step no longer happens)
Shopify wants Handle + Option1; WooCommerce wants different column names; Square rejects reordered columns.
Map columns, units, and currency
Split NT$ / JPY / KRW into Price Raw + Currency + Numeric. Convert 件入 / 箱入 / per dozen.
Save as CSV, upload via platform importer
Hope the column types match. Hope no encoding break.
Fix import errors(this step no longer happens)
Missing handles, malformed currency, variant collisions. Edit in Excel, re-export.
Re-export and retry(this step no longer happens)
Until the importer finally accepts. Half a day gone.
Convilyn workflow05 steps
Upload supplier PDF or photo
Same input. No re-keying required.
Pick target platform
Shopify, WooCommerce, or Square — schema baked in.
AI extracts, classifies, localises
Currency split, units normalised, variants resolved, language-specific patterns (件入 / 税抜 / 공급가) respected.
Download platform-ready CSV
Column names, types, and ordering already match the importer.
Direct import — no schema fix-up
Importer accepts on the first try. Half a day back.
Expected result
One supplier drop, end to end
The same catalogue, from the file the supplier actually sent to a CSV the importer accepts.
Worked example
A supplier sends a 40-page PDF price list with tax-exclusive prices in two currencies, pack sizes written three different ways, and a folder of photos named by the supplier's own article numbers.
Input
- A 40-page PDF price list, 380 products, two currencies
- Pack sizes written three different ways across the same document
- A photo folder named by supplier article number, not by your SKU
Deliverable
- A CSV in your own importer template, columns read from the file you exported
- Variants resolved against your option names, with currency split into its own column
- An exception list: products with no price, ambiguous pack sizes, and photos that matched no row
What to measure
- Time from supplier file to accepted import
- Product rows re-keyed by hand
- Importer rejections per upload attempt
- Listings corrected after publication for a wrong price or unit
System integration
Build once, call the same workflow from the web, the API or an SDK
A workflow is built once in Builder. All three call surfaces point at that one workflow, not three copies to keep in step.
In
- Direct upload
- Google Drive
- Dropbox
- URL import
- Dedicated inbox
Workflow substrate
- Excel
- Word
- JSON
Out
Call surfaces
Web interface, REST · Python SDK · TypeScript SDK. Agent workflows stream events over WebSocket; deterministic conversion is polled.
Where it runs
Workflows run in the cloud. Ainalyn for desktop (beta) handles image, PDF and OCR conversion on the machine; convilyn.local converts offline with no key and no network.
Your own tools
You can host a tool server and have the platform call into it. Every outbound request carries a HMAC-SHA256 signature and a timestamp, so you can verify it came from us.
URL import accepts https only, does not follow redirects, refuses addresses that resolve to internal ranges, and has a size limit. Download links are short-lived presigned URLs and can be reissued.
Why Convilyn
The substrate is ours, and the guarantees are checkable
Document automation projects stall at security review. Every item in this section can be checked before you talk to us.
Decision boundary
Eight gates decide in code, against fixed thresholds, in a fixed order. A model cannot skip one, and cannot argue a threshold up.
- RedactionSensitive content is masked before it reaches a model.
- Budget ceilingA per-request spend cap. Work stops at the limit.
- Retry policyFixed backoff, so a rate limit never becomes a stampede.
- Phase legalityA tool runs only in the phase the workflow allows.
- Cycle detectionThe same call repeated in a row aborts the run.
- Output reject limitThree rejected outputs and the run stops retrying.
- Tool permissionWriting to an outside system needs granted authority.
- Code rescue admissionWhether a code rescue may run is configuration, not a judgement.
Extraction is held to the same kind of rule: a field the source does not contain is left empty. A grounded blank is the correct answer; an invented value is not.
Privacy is the default, not a setting to find
Model providers do not train on your content. Runs started with an API key are excluded from Convilyn's own model work in code, and fail closed — a refusal by default, not a policy statement.
Integration flexibility, because the constraint is usually on your side
The interfaces are listed above. The other half of that commitment is this: if a requirement is beyond us, we say so during the assessment rather than at go-live.
Model training
Inputs and outputs sent through AWS Bedrock are not used to train Anthropic's models (see Bedrock's data protection terms). Runs started with an API key are excluded from Convilyn's training data, enforced in code.
Deletion
Files can be deleted immediately through the API, by the uploader only. A file attached to a running workflow is protected until that run reaches a terminal state.
Encryption and isolation
TLS in transit, encrypted at rest. Storage paths are isolated per account and every read and delete checks ownership — guessing an identifier does not reach another account's file.
Redaction
Rules match on field name and value shape, so content inside fields like resume_text is caught too. It does not rely on a model to judge what is sensitive.
Sub-processors
Every sub-processor handling customer content is listed publicly, with its role and processing region. That page is the current list at all times, and the address for questions about it is on the page.
Retention
- 1 hourSource files and download links
- 7 daysWorkflow run state
- 30 daysDeliverables
The cleanup pass runs every 15 minutes, which is why the shortest horizon is about an hour rather than exactly one. A file still attached to a running workflow is kept until that run finishes. Any file can be deleted immediately through the API.
Processing regions are fixed today: file storage and workflow execution in Tokyo (ap-northeast-1), model inference in us-east-1, and OCR in Singapore (ap-southeast-1). The region cannot be selected per request. Raise a data-residency requirement during the assessment and we will say plainly what is and is not possible.
Common questions
Questions that come up before a rollout
How long does a rollout take?
It scales with how many supplier formats are in scope rather than with how many products. One supplier's layout is usually working within the first delivery pass; each additional format is an incremental step on top of a workflow that already runs.
Do we have to change our storefront or seller centre?
No. The workflow reads the importer template you export from your own seller centre and fills its real columns, so nothing about the marketplace layout is assumed. Results come back as a file you upload, or over the REST API and SDKs if you want the step automated.
Will supplier pricing be used to train a model?
No. Supplier quotes and cost data never become training data — not for the model provider, and not for us.
What happens when a price or unit is read wrong?
Prices and units are transcribed from the source rather than recalculated, and anything ambiguous is listed as an exception instead of being guessed. A product with no readable price is reported, never published with a made-up one.
How do we start?
One real supplier file, plus the import template exported from your seller centre.
Start with one supplier file
The assessment comes back as a written scope statement naming the fields, the currency and unit rules, and the price approval point. No charge.
Book an adoption assessment