PRODUCTS
SalesField SalesCRMRewardsClaimsInventoryProcurementWarehouse ManagementField ServiceServiceSupportTelephony & MessagingFinance & AccountingPayrollExpense ManagementCommercePortalsAnalytics & ReportingData StudioMobile AppSee all products →PLATFORM
Platform overviewApp BuilderAutomationIntegrationsSecurity & GovernanceChange ManagementDevelopersSENSE AI
Sense AI overviewSense AssistSense ControlSense VisionAI StudioTrust & governanceIn Claude & ChatGPTUse casesSOLUTIONS
FMCG & DistributionManufacturing & Dealer NetworksAutomotive & DealershipsPharma & HealthcareConsumer DurablesAgri-InputsBuilding MaterialsService NetworksWarehousing & 3PLFinancial AccountingERP SoftwareIndia GST ComplianceUAE VAT & e-InvoicingSaudi ZATCA & VATAll solutions →01 · THE ASSISTANT
Ask in plain language. Answer from your own fields.
Grounded in your entities, your records, your documents — with citations. Sense runs your reports and dashboards, draws the chart when a chart is the answer, and drafts what does not exist yet; you approve it. On the web and in the field app.
Senseacting as Ravi K.Which outlets on today's beat are due a reorder?
4 of 11 — Apex Mart, Shree Ganesh, NewLeaf, Balaji Kirana. Apex is also due the scheme pitch.
Add reminders → approve
from your visit + order history
02.4 · THE SENSE FAMILY
One badge. Four ways to wear it.
Sense AssistAsk your business anything. Approve what it does next.Report-first answers · nudges · the daily briefExplore Sense Assist →
Sense ControlThe copilot that configures the system — with your permissions.Preview-first · approval-gated · 22 task recipesExplore Sense Control →
Sense VisionPhoto in, measured answer out — self-hosted vision.Shelf analysis is the first Sense Vision applicationExplore Sense Vision →
AI StudioBuild your own agents. Bring your own models.Templates · BYO keys · your own tools & capabilitiesExplore AI Studio →
02.5 · VISION
Sense can look, too.
A photo becomes a measurement
A rep’s shelf photo becomes a measured compliance score and share-of-shelf — offline capture included.
Computed, never eyeballed
Answers come from a vision pipeline, never estimated by a language model.
New SKUs taught with photos
Reference photos, not retraining — add a product with 5–10 shots and reindex.
Compliance 84%Share of shelf 62%model v2.4.1 recorded
03 · GOVERNED CONSUMPTION
Budgets that stop overruns before the spend.
Seats, credit ledgers, per-model rate cards, and hard budgets — enforced up front, not discovered on the invoice. Bring your own key if you'd rather meter it yourself.
AI usage — July
42 seats · 3 agent workspaces
Credits used64,120 / 100,000
ASSISTANT
41,300
AGENTS
18,940
DOCUMENT Q&A
3,880
Budget guard: requests pause at 100% — no surprise overage. BYO-key active for 1 model.
04 · AI STUDIO
Build agents on your model, not around it.
Agents, knowledge bases, and tools composed in a studio — multi-provider models with failover, and the open MCP standard so external AI clients connect securely to your data.
Agents
Collections chaser, beat rebalancer, claims triager — scoped, approved, audited.
Knowledge bases
Agreements, price lists, SOPs — grounded answers with citations.
Multi-provider models
Pick per task, fail over automatically, meter per rate card.
Open MCP standard
Claude, ChatGPT and Claude Code connect over one endpoint — same permissions, same audit. How it connects →
Your own capabilities
Bundle your script tools and connected tools under a name and grant it to an agent — like a permission set for AI.
Nudges
Deterministic signals on records — overdue receivables, PO quantity drift, dormant customers — that explain themselves and act with one tap.
05 · REPORTS, DASHBOARDS & CHARTS
Ask the question. Get the report, the dashboard, or the chart.
An answer is not always a paragraph. Sense runs the saved reports and dashboards your business already trusts, and when the question is about a population it draws the chart on the spot — from your live records, as you, with the working shown.
REPORT-FIRST, DASHBOARD-FIRST
A canned business question usually has an official answer already. Sense looks for the saved report or dashboard first and runs it; only when none exists does it assemble the answer from raw records.
CHARTS ON THE FLY
Aggregations come back as dashboard charts — any grouped chart kind, more than one series, period-over-period in one step. Top groups plus an honest Others bar; the true group count is always stated.
YOUR DASHBOARDS, RUN FOR YOU
Ask for a dashboard and Sense runs it through the same evaluator the dashboard page uses, as you, and returns its widgets as cards you can drill into.
SAVED REPORTS
The number the business agreed on
Sense lists and runs your saved reports under your permissions. For a canned question it prefers the report over reconstructing the logic from tables — so the figure it gives is the one finance already signs.
DASHBOARDS
Executed, not described
Ask for a dashboard and it runs — the same evaluator, filters and permissions as the dashboard page — and its widgets return as cards. Drill a tile and you are on the records it was built from.
CHARTS
Drawn the moment you ask
Any grouped chart kind your dashboards support, several series side by side, this period against last with gaps filled in. Names instead of identifiers, the true group count, the top groups plus Others. Kept when the conversation is reopened — on web and on mobile.
DRAFTS
What does not exist yet
Describe a report or dashboard that is not there and Sense drafts it in the designer’s own format. The draft lands as an approval; nothing is published until you say so.
EVERY SURFACEThe same report, dashboard and chart tools serve the chat in xMatix, the assistant in the field app, and Claude or ChatGPT connected through the MCP endpoint — under the same permissions.Why the chart is the answer →
06 · WHERE YOU ALREADY WORK
Sense, inside Claude and ChatGPT.
Connect the assistant your people already use to your xMatix organisation over one open endpoint. It asks your records, runs your reports and dashboards, searches your knowledge and — only if you allow it — acts. As the signed-in user, inside their permissions, with every call audited. Off by default; one switch, one ceiling, one log for IT.
How the connector works →CLAUDE
Ask your business from the draft
Add the endpoint as a connector; Claude signs you in as yourself and cites the records it used.
CHATGPT
Live data, not a pasted export
Same endpoint, same permissions — usable in deep-research mode through the search and fetch tools.
CLAUDE CODE
Administer from the terminal
Hand the Setup Copilot a task, approve each step, never open a browser.
IT
A switch, a ceiling, a log
Off for every organisation until an administrator turns it on. Writes are a separate switch. Every session is listed, and can be cut.
SETUP COPILOT
Admins configure by chatting — with approval gates.
"Add a 'preferred delivery window' field to outlets and put it on the mobile visit form." Sense drafts the change set; you review and promote.
07 · USE CASES
Agentic AI, stated as jobs.
Every use case as a person, their problems, the flow the agent runs, and where it stops for a human. Twelve of them, across claims, service, sales, inventory, finance, field sales, merchandising and administration.
CLAIMS
The autonomous claim validator
Claims assembled from evidence, validated against policy, triaged by Sense, approved in bulk.
SERVICE
Service opportunity discovery
Contracts and schedules read overnight; the call list is ranked and ready at open of business.
SERVICE
The engineer’s assistant
Asset history, the knowledge base and a drafted parts list, on the job card, in the field.
INVENTORY
The parts manager’s reorder
Three replenishment engines, one explainable line, and a nudge that catches the covered PO.
FINANCE
The collections chaser
Receivables aged as they happen, reminders drafted per account, every one waiting for approval.
CRM
The AI-assisted contact centre
The caller matched before the answer, routed by skills and presence, summarised after.
Questions the AI-curious ask.
Is our data used to train models?+
What happens when a user lacks access to the answer?+
How are credits priced and metered?+
Can we bring our own model key?+
What can agents do without a human in the loop?+