The Platform [CXM]

The whole loop, screen by screen.

Syna Lens, Collect, Sense, Monitor, Action, and the governance underneath. Shown working, in order.

Click a stage to jump to it
Sense tags in Monitor tables Syna Lens the AI analyst, answers in plain words Collect captures feedback, every channel Interpret classifies open feedback Monitor analyzes, down to each journey Action closes the loop, on the record Admin: roles and residency, under everything.

SYNA LENS

Ask a hard question. Watch it dig.

Ask in plain language. Syna Lens investigates the way an analyst would, tracing a score down to the drivers, the issues, and the actual comments, deciding its own next step as it goes.

Illustrative demo
Syna Lens
AI analyst
Ask a question about any survey, program, or segment
Watch it work

COLLECT

The listening layer.

Classic or conversational surveys, multilingual and on-brand, over email, SMS, and WhatsApp. On your infrastructure or ours.

collect.numr.app
Meridian Onboarding · Edit
EN
Publish
QUESTIONS
DESIGN
+ Welcome Screen
Q1. Based on your recent account opening experience with Meridian, how likely are you to recommend us to friends, family, or colleagues? Rating
0to10
012345678910
Choose a question type
Rating
Text
Single Choice
Multiple Choice
Like / Dislike
Number
Date / Time
Q2. {{pipe}} logic We are sorry your experience did not meet expectations. Tell us how Meridian could improve.
Q3. Which areas most influenced your account opening experience?
Meridian
Based on your recent account opening experience with Meridian, how likely are you to recommend us?
Not at all LikelyExtremely Likely
powered by numr
Live preview
Distribution · Invites
Save
Initial Invite
Email
SMS
WhatsApp
Invitation Text
Subject
Your feedback matters to us
Dear {{first name}},
As a valued customer we would like to invite you to share your feedback.
{{SurveyURL}}
Send
Immediately
Access · Delivery
Send Surveys Between
Any day, in selected hours
01234567891011121314151617181920212223
Survey Send Rules
Do not send surveys if
Surveyed in last
30
days
Illustrative data
Build
Classic and conversational surveys
Question types: rating, text, single and multiple choice, like/dislike, number, date/time, 0-10 recommend scale, info screens
Configurable scales: 5, 7, 11-point, or any
Visualizations: linear, radial dial (dial on mobile), or icon-only with swappable icons
Per-question properties: steps, shape, color scale, labels, anchor text and icons, mandatory
Auto-progress to the next question
Logic: skip, jump, hide, conditional display
Piped and dynamic text
Welcome screens and sub-questions
Brand
20+ preset themes
Full customization: fonts, typography, colors, header and footer, opacity
Custom logo
Your own survey domain, anonymous links, QR codes
Track and integrate
Per-channel delivery and response status
Link expiry and close dates
Bulk actions: resend, re-expire, export
Send and export webhooks
Multilingual: per-language translation, add languages, import/export
Send
Email, SMS, WhatsApp
Initial invites and reminders
Send now, scheduled, or within set hours
Send rules so no customer is over-surveyed
Bring your own sending infrastructure
Custom from, reply-to, unsubscribe

SENSE

Your taxonomy. Your rules, on the record.

Define the categories, train the model by hand, test it before it goes live, and version every deployment. The AI answers to you.

ai.numr.app
Meridian Onboarding Model · Edit
BUILD MODEL
FINE TUNE
TEST
LOG
Build Actions
Set Context
See Prompt
Minimum 25 characters
Show Code
Name
Number of comments in the category
Active
Onboarding experience
1,284
Account opening
742
Document upload
318
Identity verification
205
Approval turnaround
141
+ Add New L3 Category
Branch onboarding
96
Welcome communication
74
+ Add New L2 Category
Digital banking setup
863
Relationship manager
388
Generic
47
+ Add New L1 Category
BUILD MODEL FINE TUNE TEST LOG
Search for word, phrase or ID
100 respondents
Run
Classification strength
Agreement with the set you labeled by hand
92%
184 of 200 hand-labeled comments matchedStrong
CategoriesFrequency
Onboarding experience
38%
Digital banking setup
24%
Card services
16%
Relationship manager
14%
BUILD MODEL FINE TUNE TEST LOG
Set Variable
200 comments
Strong
Model strength
Comment 3 / 200
The document upload step kept failing and I had to restart my application twice before it went through.
Classify and Score
Onboarding experience Account opening Document upload
Sentiment Negative
Reassign
Approve
Illustrative data
Define
Build an L1 / L2 / L3 taxonomy
Add, reorder, and toggle categories
Set context and prompt per node
Auto-generate or build by hand
Stable code numbers so trends hold when categories change
Version and analyze
Versioned deployment log
Reprocess history so trendlines survive
Publish anytime
Sentiment and cause/topic classification
Works on comments, closing remarks, and chat
Models shareable across projects
Train
Human-in-the-loop labeling
Model-strength indicator
Classify and score
Test
Test before deploy
Category frequencies
Search by word, phrase, or ID

MONITOR

Every stage, priced to the next outcome.

Monitor quantifies how the experience at each stage changes the odds a customer returns, renews, or buys again. Built by your analysts, gated by significance.

monitor.numr.app
Meridian · Onboarding health Draft
Last 3 months
+ Add widget
Widgets
KPI Gauge
Sentiment
Priority Box
Key Drivers
VOC
NLP
Table
Drag onto the canvas to build any view.
Key drivers · Priority matrix Regression-backed
PRIORITIZE LEVERAGE IMPROVE MAINTAIN Onboarding speed App reliability Fee clarity PERFORMANCE → IMPACT →
Renewal drivers by segment Correlation
CORRELATION WITH RENEWAL (r) Mass affluent 0.68 Priority 0.52 Digital-first 0.60 Wealth Base too low, suppressed (n=14) —
95% confidence · bases under 30 hidden automatically
Format · Key drivers
Chart type
Priority matrix
X axis label
Performance
Y axis label
Impact
Statistics
Suppress low bases
Confidence threshold
90%
95%
99%
Auto-hide below threshold
Customer journey
CUST ••• 3821
Account opened
Day 0
Onboarding survey · completed
Day 3
Mobile app first login
Day 9
Support call · 6 min
Week 6
Servicing survey · completed
Month 4
Relationship review booked
Month 5
Survey event Non-survey event
Journey Outcome
Servicing → next service
High impact
10x
more likely to return
for the next service
Customers whose servicing experience scored in the top tier, versus the baseline experience.
n = 2,940 · masked IDs Change in odds, not a revenue guarantee
Illustrative data
Analyze
Correlations and regressions out of the box
Regression-backed key-driver priority matrix
Significance testing and low-base suppression
Auto-hide anything below your confidence threshold
Performance indices
Segment, geo, and journey analysis
Sentiment and VOC from Sense
Build
DIY dashboards
Widget palette: KPI gauge, sentiment, priority box, key drivers, VOC, NLP, tables
Excel-grade formatting
Computed variables, nets, calculated expressions
Per-option filters
Impact
Quantify how each stage changes the odds of the next outcome
High, mid, low impact tags per stage
Tie experience to return, renew, or repurchase (always more likely, never revenue proof)
Journey and explore
Per-customer journey timelines (survey and non-survey events)
Respondent grid with Sense classification tags
Filter across many dimensions
Export

ACTION

Fix the issue. Keep the receipts.

Every ticket gets an owner, a route, and a clock. Every move becomes proof you can point to.

action.alert.app
Cust ••• 4821
Open:Jun 30, 11:20 AM
Time Overdue: Jul 3, 11:20 AM
Status
Waiting On Us
Assigned To
priya.n@meridian.io
Notes
Email
Call
Upload
Resurvey
Ticket Details
Select Activity
Select Members
P
Classify L1
Option
L2
Option
Apply
Activity Customer Journey
Today
Alert Status Wed Jul 08 2026
Note: Waiting for IT
priya.n@meridian.io moved ticket from New to Waiting On Us.
Status Change Note Wed Jul 08 2026
Waiting for IT
priya.n@meridian.io left a note.
Jul 1
Assignment Tue Jul 01 2026
arjun.r@meridian.io assigned ticket to priya.n@meridian.io.
Detractor Alerts
Open 33 Overdue 12
Grid List
Search
NEW 25
Cust ••• 3907
Due Jul 2
A Overdue
Cust ••• 5142
Due Jul 4
D Overdue
ON US 1
Cust ••• 4821
Due Jul 3
P Overdue
ON CUST 6
Cust ••• 2288
Due Jul 6
R
CLOSED 2
Cust ••• 1760
Closed Jul 5
L
Configure statuses
STATUS NAME OPEN/CLOSED NOTE ASSIGN
New Open
Waiting On Us Open
Waiting On Customer Closed
Closed Closed
+ Add new status
Illustrative data
Board
Alert board with kanban statuses
Grid or list view
Open and Overdue counts
Filters across many dimensions
Drag tickets across statuses
Different alert rules create different boards
Prove
Full activity and audit log per ticket
Who did what, and when
Configurable statuses: open or closed, require a change note, control assignment and membership
Every close recorded on the ticket
Own and route
Assign an owner to every ticket
Route by rule
SLA and overdue clocks
Assignee avatars
Nothing left unowned
Act
Act on the customer from the ticket: note, email, call, upload, resurvey
Reclassify against the Sense taxonomy (L1, L2, L3)
Status-change notes

ADMIN

Locked down. Not weighed down.

Role-based access, SSO, and governed integrations for every deployment. When regulation requires it, optional on-premise tokenization keeps PII inside your firewall.

admin.numr.app
N Admin
ACCESS
Roles
Groups
Users
SSO Connected
Okta SAML · enforced
Licenses 42 / 60
28 Creators14 Explorers
Roles & Access
Meridian Bank workspace
Create New Role
Creators Explorers
Creators have full access. Roles and groups scope what Explorers can see and do.
ROLEMEMBERSDATA ACCESSRESTRICTIONS
Regional Analyst 8 Retail programs Export off
Exec Viewer 6 All dashboards Read only
Program Manager 12 Own programs only Masked contact
Insights Analyst 15 Assigned dashboards Read only · PII masked
Branch Lead 4 Own programs only Per-user scope
Per-user restrictions and group membership override role defaults.
Illustrative data
OPTIONAL · ON-PREMISE RESIDENCY Most teams run standard cloud. When regulation requires it:
YOUR INFRASTRUCTURE NUMR CLOUD YOUR FIREWALL customer data tokenization module structured PII tokenized at source de-tokenize API re-identify locally local dashboards re-identified locally delivery processes tokens only free-text redaction PII stripped on ingest analysis processes tokens only Sense classification our classification and tuning layer processes tokens only tokens only free-text redacted on ingest results, tokenized Structured PII tokenized at source · Free-text redacted on ingest · Re-identification only inside your firewall. For the teams that need it. Optional on-premise tokenization keepsstructured PII inside your firewall; onlytokens cross to our cloud.
Access and roles
Single sign-on (SSO)
Role-based access control
Create custom roles
Per-user restrictions
Groups
Creator and explorer roles
License management
Integrations
Out-of-the-box connectors: Salesforce, Oracle, Freshdesk, Twilio, Gupshup, and more
Send and export webhooks
Webhooks built for you by our team, out of the box
Governed inbound and outbound integration
Optional on-premise residency
On-premise tokenization module
Structured PII tokenized at source
Cloud processes tokens only
Free-text redacted on ingest
Re-identification only inside your firewall
Run on your infrastructure or ours

You have seen the tour. Now bring your data.

A Numr demo runs on your questions, not a slideshow. Pick a customer and we open the journey. Ask a hard question and Syna Lens digs. If a platform cannot do this live, you have learned something useful too.