The Aiden walkthrough of MAIA is the cleanest end-to-end demo we have on the AI assistant inside BNTouch. In about seven minutes, Aiden runs three live tasks against a test borrower record: he asks MAIA to list closed mortgage records this month with a dollar total, he creates a follow-up task tied to a specific borrower at a specific time, and he activates a new loan follow-up campaign on a funded file. Three jobs, one chat window, no menu hunting. The video is worth watching twice if you have ever lost a morning to clicking around the CRM.
Aiden runs product demos for the BNTouch team. The reason this walkthrough lands harder than a static feature page is that he uses MAIA the way a working loan officer would use it on a Wednesday afternoon. A pipeline to triage, a borrower to follow up with, a campaign to activate. He does all three back to back without leaving the assistant pane.
The framing he opens with is honest about scope: “[MAIA] is going to have an amount of tasks that she can do, and again, we’re gonna roll those out over time” [4aoHsOOJ3Z0 m1]. Translation: this is not a fully autonomous agent. It is a chat layer over the CRM that handles a specific list of jobs reliably today, with more coming.
Task one: pull closed loans this month with a dollar total
Aiden types a request for closed mortgage records this month. MAIA returns the records and the total volume number. From the demo: “She gave us before, mortgage records that were closed this month, and then with the total amount so you can see how much was closed this month in the numerical value” [4aoHsOOJ3Z0 m3].
The point here is not that the CRM can run that report; the reports tab can do it too. The point is that MAIA collapses what would be three clicks and a date range filter into one sentence typed in the morning. Branch managers running daily standups will recognize the value immediately. A loan officer who wants to know their own number before a 1:1 will also recognize it.
Task two: create a task tied to a specific borrower at a specific time
Aiden uses the test borrower Rusty Shackleford for the task demo. He types “call Rusty Shackleford” and starts the create-task action: “We will do, create task, right? That’s the first thing that we want, the action we want to happen. Call Rusty Shackleford on Wednesday at 2 p.m.” [4aoHsOOJ3Z0 m4].
MAIA confirms back what she heard, sets the task on the right borrower record, and writes it into the calendar for Wednesday afternoon. The detail worth noting is how she handles relative dates. “Today, if we said, you know, call Rusty on Monday at 2 p.m., since Monday already happened this week, it’ll actually set for the next coming Monday. So it always sets for the next coming day that you have available” [4aoHsOOJ3Z0 m5]. The behavior matches what most people expect. If you ask for Monday and it is Thursday, the task goes on next Monday, not the Monday that already passed.
Task three: activate a new loan follow-up campaign on a funded file
The third demo is the one most worth copying as a daily habit. Aiden pulls up a recently funded loan, references it by loan number, and activates a new loan follow-up campaign on that record. From the demo: “We’ve just gone through, had [MAIA] pull the information of all of our new funded loans. [MAIA] then created a task to call them the following day at 2 p.m., and then we [had her start the new loan follow-up]” [4aoHsOOJ3Z0 m7].
The reason this matters: post-funding follow-up is where loan officers leak future refis and referrals. Borrowers feel taken care of for 30 days, then the LO goes quiet, and four years later the borrower refinances with someone else. A daily MAIA habit that pulls every funded file from the last week and activates the post-funding campaign closes that gap.
Useful detail: reference by loan number, not borrower name
Halfway through, Aiden mentions a workflow most demos skip. “[Use the] loan number because you can copy and paste it right out of there. It’s even easier than typing the name of the [person] or the loan file name” [4aoHsOOJ3Z0 m6]. If you have two borrowers named John Smith, the loan number disambiguates instantly. If you are looking at a record in another tab and want MAIA to act on it, the loan number is the cleanest handle.
What this demo does not show (and you should know about)
Three honest limits that the Aiden video does not cover, but which matter when you calibrate expectations:
MAIA does not place outbound calls. The BNTouch team has stated this on other demo videos: there is no AI voice agent. She drafts the task and the message, you still pick up the phone.
MAIA does not yet send SMS or build campaigns from scratch in this demo. Aiden mentions these as roadmap: “Eventually she will be able to, you know, send text messages for you, create new campaigns” [4aoHsOOJ3Z0 m1]. Today she activates existing campaigns and creates tasks. Drafting outbound SMS lives in the email editor and the campaign step builder, where MAIA pulls in ChatGPT for content help.
This is a search-and-retrieve plus task assistant in this video. The newer NextStep recommendations capability covered elsewhere on the channel is not in this clip.
How to use this demo as a 15-minute team training
If you have a branch of three to ten loan officers and you want them all using MAIA by next Friday, the simplest play is to gather the team, play the Aiden walkthrough, and assign each person the same three tasks against their own pipeline on the spot. Pull your closed loans this month. Create one task on a real borrower. Activate one campaign on a recently funded file. Twenty minutes of live practice with the assistant beats an hour of reading documentation.
Try MAIA on your own pipeline
If you want to run the same three tasks Aiden ran but against your own records, request a demo and ask the BNTouch team to enable MAIA on a sandbox version of your data. The full capability list is on the MAIA AI page.
