Session Notes: Karen — Product Entry
June 3, 2026 · Nate St. Pierre, Karen
Overview
This session focused on how products get added and maintained across RCB's three systems: Antera Advance (the primary product database), Sage (the distributor platform that feeds the EI website), and ASI (a second distributor platform). Karen walked through the current workflow using the crystal products as a working example, and the conversation surfaced both the catalog gap and the manual effort required to keep all three systems in sync.
Who Adds Products and Why
Product entry into Antera happens for several different reasons, and different people handle each:
Cindi or Briana add products when RCB is expanding its catalog — deciding to offer a new product line or adding styles that haven't been sold before.
Sales adds products on the fly during customer calls, when a customer wants something that exists (and RCB can make or source it) but hasn't been entered into Antera yet. This happens regularly — a customer might request a plate size that's technically available but not yet in the system. Sales enters what they need to process the order and moves on, without following a standardized process.
Karen adds parts — components that are part of a bill of materials for a finished product — rather than catalog products. Her additions are more structured and methodical.
The result is that product entries vary widely in quality and completeness depending on who entered them and under what circumstances.
How Product Entry Actually Works
The standard approach is to clone an existing product in Antera and change the relevant details. Starting from scratch is rare — most new products are similar enough to something already in the system that cloning is faster and produces a more complete record.
There is a checklist for adding new products that specifies which fields need to be filled in and how. Karen mentioned that she doesn't always use it herself, and sales entering products during customer calls certainly doesn't follow it. For large batches of similar products, an Excel import is faster than entering records one at a time.
The result of inconsistent process: Karen has spent significant time combing through Antera correcting entries — wrong field values, missing required fields, pricing that was entered for a single order and never validated as a general price. She described this as an ongoing cleanup effort.
The Catalog Gap: Crystal Products
Cindi recently came to Karen about the crystal product catalog. Currently, 34 crystal products are listed on the EI distributor website (which is driven by Sage). The actual universe of available crystal products is much larger — hundreds of options exist — and many of them are already in Antera from previous orders. They just haven't been pushed to Sage and therefore don't appear on the website.
The constraint is time: pushing products from Antera to Sage requires exporting from Antera, reformatting the data to match Sage's import template, and then running the import. There's no one dedicated to doing this, so it doesn't get done. The catalog has been static at 34 crystal products while potentially dozens or hundreds more could be listed.
How Data Gets from Antera to Sage and ASI
Antera does have a built-in push to Sage, but Karen's experience is that it doesn't reliably populate all the required fields — you get a partial product record in Sage that still needs manual work. So the native push isn't practical as a workflow.
Karen's current process: 1. Export the relevant product data from Antera as an Excel file 2. Use Claude to remap the data to Sage's import format — the column names and structure are different between the two systems 3. Import the transformed sheet into Sage
ASI is the same story with a different format — a separate import, separate column structure, done independently.
Karen does this process herself. It's not documented, and Cindi uses a different approach (manual entry, or a different variation of the same chain). The workflow works but it's fragile: Karen is the only person who does it this way, and the Claude-assisted transformation produces slightly different results each time (she flagged this explicitly — she understands that AI models aren't deterministic and produces consistent-enough results for her purposes, but it's not a reliable shared process).
A Better Architecture: Single Input, Multiple Outputs
Nate proposed an alternative during the conversation: rather than running export → transform → import chains per system, build a single AI-powered intake where product data goes in once and the system generates correctly formatted import files for all three systems simultaneously.
Karen responded positively to this ("that's beautiful") and started stress-testing it herself. Her main concern: she currently uses Antera as the internal-first check — she enters into Antera before pushing anywhere else, to make sure the company's system of record is always accurate before anything appears on the distributor websites or the retail page. A single-input approach would need to maintain that same discipline.
The refined version they landed on: Antera stays the starting point and source of truth. The change is what happens after — instead of manually exporting and re-importing for each downstream system, an AI-assisted step generates the Sage and ASI import files automatically from the Antera record. One transformation step instead of three manual chains.
The same logic applies to pricing updates, which follow the same Antera-first → push downstream pattern.
Antera API
Nate noted that Briana had a meeting with the Antera rep recently, and he had asked her to inquire about API documentation. If the API exposes read and write operations, it could enable a more automated bridge — rather than running Excel import/export chains, the system could push directly to Sage and ASI when a product is created or updated in Antera. The answer from the Antera rep is still pending.
Process Standardization
Nate noted at the end of the conversation that regardless of how the technical workflow evolves, having a documented and agreed-upon process for product entry — accessible to everyone who does it — would be valuable on its own. Right now, different people do it differently, and there's no shared reference. Standardizing the process (even without any AI component) would reduce the data integrity problems Karen is spending time cleaning up.
Follow-Ups
- Get the Antera API documentation from Briana — this determines which automated options are feasible
- Ask Karen to share the existing product entry checklist
- Get the Sage and ASI Excel import templates — these define the field schemas needed for the transformer
- Ask Karen (or Briana) how many products are currently in Antera but not on Sage — this sizes the crystal catalog expansion opportunity