ElderMark does not expose a public API and is deliberately restrictive about external data access. This makes it impossible to build real-time integrations or trigger external automations (e.g., email nurture sequences in MailChimp) directly from ElderMark events.
The workaround is to automate the manual export process using a browser automation tool, so that contact data can be extracted on a schedule and fed into downstream systems without requiring manual effort from the client.
Affects: [1] — see [2] for full context.
Axiom.ai is a Chrome extension that records a sequence of browser interactions (clicks, text entry, navigation) and replays them on a schedule. It requires no code and can handle login flows, navigation, and file downloads.
Pricing note: The starter tier (~$30–50/month) supports one bot and is sufficient for a single scheduled export. Factor this into client billing.
Key limitation: Axiom.ai bots are brittle — if ElderMark changes any UI element (button label, position, DOM structure), the bot will silently fail or error out at that step. Plan for periodic maintenance.
The goal is to automate a weekly export of the full "All Prospects" contact list as an XLS file.
ElderMark's login page has a redirect quirk: navigating to the root URL causes a redirect that confuses the bot's field-targeting. The fix is to start the bot at the direct tenant URL (e.g., app4.eldermark.com/Anthemian/) rather than the generic login page.
The login form has three fields:
1. Company Code — labeled "login" in the help text (confusingly)
2. Username — the actual user login
3. Password
Steps:
1. Navigate to app4.eldermark.com/[tenant]/ (direct tenant URL)
2. Enter text → Company Code field
3. Enter text → Username field
4. Enter text → Password field
5. Click → Login button
Credentials are stored in LastPass.
After login, ElderMark may land on different views depending on the user's last session state. To ensure consistent navigation regardless of landing page:
Note: Steps 2–4 are intentionally redundant to handle variable post-login states. The bot navigates through the full menu path every time rather than assuming a starting view.
Once on the All Prospects list:
The file downloads directly to the local machine. It does not email the file — a subsequent automation step (e.g., a folder watch or upload script) would be needed to move it to a shared location.
Blocker: The bot consistently fails at the "All Prospects" click step.
The element is visually identified and highlighted during the run, but Axiom.ai reports "couldn't find element: All Prospects" and errors out. This has occurred across multiple bot configurations and navigation path variations.
Attempted mitigations (unsuccessful):
- Added millisecond delays between steps
- Tried alternate navigation paths to reach the All Prospects view
- Rebuilt the bot from scratch multiple times
Hypothesis: The All Prospects link may be rendered in a way that Axiom.ai's element selector cannot reliably target (e.g., dynamically injected, inside a frame, or using non-standard click event handling). ElderMark may also have bot-detection that interferes after the login step.
Next action (owner: Mark Hope): Resume troubleshooting with a fresh session; investigate whether the element can be targeted by XPath or CSS selector rather than visual click recording.
If the Axiom.ai bot cannot be made reliable, the fallback is to ask the client (Erica) to perform the export manually on a weekly cadence and drop the file in a designated shared folder.
eldermark-prospects-YYYY-MM-DD.xls) so downstream automations can reliably locate the latest file.This is the recommended interim approach while the bot issue is investigated.